A BALANCE SHEET FOR THE NATION:
A STUDY IN CONCEPTS
JOHN 0. BLACKBURN
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
The writer acknowledges his indebtedness to Dr. John N.
Webb, chairman, and the other members of his supervisory
committee, Dr. Clement H. Donovan, Dr. James S. Lanham,
Dr. C. Arnold Mlatthews, and Dr. Cecil G. Phipps, for their
helpful comments and criticisms.
He is also grateful to Graeme S. Dorrance of the Research
and Statistics Department, International Monetary Fund;
Raymond W. Goldsmith of the National Bureau of Economic Research;
John Gorman of the National Income Division, Department of
Conmerce; and Stephen P. Taylor of the Division of Research and
Statistics, Board of Governors of the Federal Reserve System,
for their making time in busy schedules to discuss the project.
Appreciation is expressed to the College of Business Admin-
istration of the University of Florida for assistance rendered in
the form of teaching assistantships in 1956, 1957, and 1958, and
a fellowship awarded in the fall of 1956.
Appreciation is also expressed to Mrs. Ada Kip who typed the
final manuscript, and to Mrs. Lewis F. Haines and Mrs. Jimmy C.
Perkins of the Graduate School for their helpfulness in answer-
ing many of the questions which arose in preparing the manuscript
in its final form.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS . . . . . . . . ..
LIST OF TABLES . . . . . . . . . .
INTRODUCTION . . . . . . . . . .
I. THE SCOPE OF NATIONAL AND SECTOR BALANCE
SHEETS . . . . . . .
II. VALUING SECTOR AND NATIONAL WEALTH ..
III. SECTORING THE ECONOMY . . . . .
IV. SAMPLE NATIONAL BALANCE SHEETS FOR
1949 AND 1950. . . . . .
V. PROBLEMS IN CONSOLIDATION . . . .
VI. THE NATIONAL BALANCE SHEET, THE NATIONAL
INCOME, AND THE FLOW OF FUNDS . .
VII. SOME IMPLICATIONS FOR ECONOMIC AND
MONETARY THEORY . . . . .
CONCLUSION . . . . ........ . .
BIBLIOGRAPHY . . . . . . ..... .
LIST OF TABLES
1. Returns, Unrecovered Cost, and Depreciation
of a Hypothetical Asset . . . .. 40
2. Comparative Methods of Sectoring the
Economy in Various Flow-of-Funds Studies . 64
3. Proposed Sectors for the National Balance
Sheet, with Major Sector Components . . 79
4. Implicit Deflators for National Assets . 87
5. National Balance Sheet, United States of
America, End of 1949. ............. 95
6. National Balance Sheet, United States of
America, End of 1950 ............. 96
7. National Income and Product Account, 1950 . 118
8. National Income and Product Account, 1950,
Revised . . . . . . . . 119
9. Changes in National Tangible Assets, 1950 . 120
10. Gross National Product and Related Flows
by Sectors, 1950 .............. 122
11. Explanation of Adjustments in Table 10 . 123
12. Changes in National Tangible Assets by
Sector and by Asset Classification, 1950 . 125
13. Changes in Financial Assets, Liabilities,
and Net Worth by Sectors, 1950. . . . 127
14. Returns from National Wealth, 1950 . . 131
It is sometimes said that we do not know enough about the Amer-
ican economy. It might be more accurate to say that we do not have
enough information of the right kind. Certainly there is no shortage
of data--a quantity sufficiently large to fill a fair-sized library
accumulates nearly every month. What has been lacking is a structure
for organizing the massive quantities of facts into summary form so
that one individual with limited time at his disposal can find out what
he wants to know. Such an organizing device is provided by the national
income and product accounts in which literally millions of detailed bits
of data are organized into an easily digested summary statement. These
accounts are well suited to measure the output of final products and
the incomes which are generated in the production process. Other im-
portant information, however, will not fit into this particular mold
and, as a result, cannot be satisfactorily shown.
Two notable examples are flows of intermediate products and trans-
actions in financial instruments. The former are not shown since in the
income and product accounts attention is focused on final products only.
The latter group, which includes such transactions as borrowing and re-
paying, escapes attention in that neither incomes nor products are direct-
ly involved. Input-output analysis may be viewed as an attempt to supple-
ment the product accounts by furnishing information about the flow of
intermediate goods. Flow-of-funds analysis may be viewed as an attempt
to supply financial information which is of great importance in
analyzing the economy, especially with respect to the all-important
savings-investment process. The preparation and analytical use of
these various measures of economic activity are generally called social
accounting, national accounting, or economic accounting.
There is still another large body of data which as yet cannot
be accommodated within the recently expanded framework of the social
accounts--namely, balance sheet information. Assets and liabilities--
money balances, mortgage debt, consumer credit, for example--all play
vital roles in the economic affairs of the modern world, yet informa-
tion regarding these factors has for the most part remained outside
the growing and increasingly well-organized body of descriptive and
analytical data provided by the various social accounting systems.
This situation, which only in the very recent past has begun to
be corrected, is all the more surprising in view of the fact that we
are almost daily supplied with information which on closer examination
turns out to be merely a part of the national balance sheet. There are
regularly published figures for the federal government's outstanding
debt and its ownership, the money supply, bank reserves, mortgage credit,
security credit, consumer credit, investments abroad, assets of insur-
ance companies, corporate assets and liabilities, consumer assets and
liabilities, and farm assets and liabilities--to mention only the more
prominent kinds of available data. We are frequently left with the un-
fortunate impression, however, that each of these magnitudes exists in
a vacuum, when in reality each is only a part of the over-all national
structure of assets and liabilities. Through this structure the eco-
nomic life of the nation is bound together as surely as through the in-
come and expenditure transactions already summarized and published in
the income and products accounts.
This study is, then, a continuation of and a supplement to recent
work aimed at filling this large and embarrassing gap in our economic
statistics. A full summary of prior studies in this area is not attempt-
ed, though one can hardly proceed without mentioning the name of Raymond
W. Goldsmith. His work, more than that of any other person, has brought
the day closer when national balance sheet data can be prepared on a
regular and official basis. The reader is referred to his several
studies listed in the bibliography, of which his exhaustive savings study
is perhaps the most significant contribution.1 As one cf the results of
this and other works, Goldsmith has prepared national balance sheets for
benchmark dates from 1900 forward, the latest published one appearing in
the thirty-seventh annual report of the National Bureau of Economic Re-
search.2 Another of his articles contains a summary of past work in the
field of balance sheet or wealth statistics, which is one of the reasons
why such a summary is not attempted here.3 The other reason is that
1Raymond W. Goldsmith, A Study of Saving in the United States
(3 vols.; Princeton: Princeton University Press, 1955).
2National Bureau of Economic Research 37th Annual Report (New
York: National Bureau of Economic Research, 1957).
3Raymond W. Goldsmith, "Measuring National Wealth in a System of
Social Accounting," Studies in Income and Wealth, Vol. XII, by the Con-
ference on Research in Income and Wealth (New York: National Bureau of
Economic Research, 1950).
Goldsmith's work virtually eclipses anything done in the past, at least
with respect to figures for the United States.
As the reader may surmise, much of the present study covers ground
that has already been subjected to some exploration--some intensive, some
cursory. The study is primarily concerned with examining and adding to
the conceptual structure within which national assets and liabilities are
measured. It should be stated that the study is not primarily concerned
with collecting balance sheet data or with some of the baffling problems
which confront the would-be collector. As one can well imagine, there
is ample scope for several studies in this area alone. It is probably
fair to say that the task of gathering and refining the necessary balance
sheet information on a continuing basis is well beyond the resources of
an individual investigator. It is, of course, necessary to introduce
figures into the study, which figures are drawn from a variety of sources.
The writer acknowledges his complete dependence upon other investigators
for such figures as are available and his use of some rather crude
guesses for figures which are not available. The results do not pretend
to attain the degree of accuracy achieved in officially prepared and pub-
lished figures, but they do illustrate the concepts which have been
One of the recurring themes of the study is the development of an
integrated social accounting system. As the writer envisions such a sys-
tem, it would involve the preparation, on a mutually consistent and inter-
related basis, of statements for national income and product, flow-of-
funds, balance of international payments, the national balance sheet, and
input-output. There are, needless to say, some people who doubt the
wisdom of trying to integrate into one system these various approaches
to measuring macroeconomic activity. This is one controversy the
writer prefers not to enter; he admits to a bias in the opposite direct-
ion. It seems relevant to observe, however, that a fully integrated set
of accounts imposes a discipline on each of the components which has
heretofore been lacking in that each statement must relate satisfactorily
to the others.
It would seem appropriate that a writer who proposes to devote an
entire study to the development of a national balance sheet should begin
by justifying his efforts through a clear demonstration of the need for
such a statement and a description of some of its possible uses. These
steps are not taken for several reasons. One is that any study, if it
is to be held within manageable bounds, must proceed on the basis of
certain assumptions; there is simply not enough time to try to prove
everything. One of the assumptions upon which this study is built is
that a national balance sheet will provide useful information, the value
of which will more than offset the trouble and expense involved.
Another reason for the omission of a long justification is that
others have already made what to the writer is a conclusive case. As
evidence there are offered some quotations from a recent report on the
state of the nation's economic accounts and some related testimony before
a Congressional subcommittee. The report and the accompanying expert
U.S. Congress, Joint Economic Committee, Subcomaittee on Economic
Statistics, The National Economic Accounts of the United States, Hearings,
85th Cong., 1st Sess., October 29 & 30, 1957 (Washington: U.S. Government
testimony recommended, among other things, the beginning of official
work on a national balance sheet. A few relevant quotations follow:
The committee feels that as part of a long-range program
of improvement and expansion of our system of national accounts
the development of comprehensive and consistent national and
sectoral balance sheets on a regular periodic (if possible
annual) basis should be taken in hand as soon as feasible.
The coanittee,however, recognizes that there are still so
many unresolved conceptual problems in this field and that the
estimates are in many cases necessarily still so rough that the
next step should not be the immediate attempt by a Government
agency to develop balance sheets or even national wealth state-
ments. It seems to the committee that this is the field for a
thorough study, exploratory and experimental in part, possibly
by one of our private research institutions.
We are convinced that the development of a flexible inte-
grated system of national accounts comprising the national in-
come accounts, the money-flow statements, the input-output tables,
the balance of payments, and the national balance ,heet, is the
most important long-range objective in this field.
More recently it has become obvious that the flow of funds
structure is not sufficient to analyze many of these problems,
and that we are going to need information on the assets, liabilities,
and liquid positions of various sectors of the economy in order to
make a more adequate appraisal of inflationary pressures and the
problems of full employment.7
National balance sheet: In view of my long-term interest in
this relatively neglected field of social accounting, I approve
the committee's recommendation that work begin as 3oon as feasible,
Printing Office, 1957). This publication is referred to in the follow-
ing paragraphs as "Hearings."
5From the report of the National Accounts Review Committee,
National Bureau of Economic Research. Ibid., p. 256.
6From a statement by Raymond W. Goldsmith, Chairman, National
Accounts Review Committee. Ibid., p. 6.
7From a statement by Richard Ruggles, Professor of Economics at
Yale University. Ibid., p. 23.
page XIV-16. In fact, I consider this a project of greater
urgency, and raise the question, "When do we begin?"8
Before proceeding to a brief preview of the contents of the study,
it is appropriate to concentrate attention for a few moments on some of
the basic terms which appear immediately and which recur throughout.
One such is the notion of an economic unit, by which is meant any indi-
vidual or group of individuals who act as a unit in making economic de-
cisions--selling, earning, spending, saving, and the like. The family
(or one of its variants) is an economic unit, as is U.S. Steel or the
federal government. The principal economic units for the purposes of this
study are households (or families or spending units), business firms,
non-profit organizations, and governments.
Another fundamental concept is that of sectors in the economy.
Whatever the values of macroeconomic analysis (and in the writer's
opinion they are considerable) one does not proceed very far before the
need for some disaggregation becomes apparent. If disaggregation pro-
ceeds all the way down to the individual economic unit, we are back to a
microeconomic level of analysis; some intermediate level between the
nation at one extreme and the individual economic unit at the other is
needed. The Marshallian industry is one of the devices used to fill this
gap, but goes too far down the scale for most macroeconomic analysis.
The sector, then, is introduced as the first stage of disaggregation below
that level of analysis represented by national aggregates. This particu-
8From a statement with respect to the recommendations of the
National Accounts Review Committee by Martin S. Gainsbrugh, Chief Econ-
omist, National Industrial Conference Board. Ibid., p. 53.
lar aspect of analysis is discussed at length in Chapter III, though it
is of necessity introduced earlier in the study.
The following paragraphs are devoted to a brief statement of the
contents of the study so that the reader is provided with some idea of
where he is invited to go while he is on the way. The goal which the
writer has attempted to keep constantly in view is that of developing a
statement of assets and liabilities by sectors and for the nation as a
whole, a statement which can be related to the other more highly develop-
ed measures provided by national income and flow-of-funds.9 Some de-
tours are apparently unavoidable, though every effort is made to return
to the main stream of development as soon as possible.
Chapter I deals with the relationships between the various concepts
which a study on the borderline between economics and accounting must
somehow reconcile--namely, wealth, assets, and capital. A working set
of definitions is adopted for the purposes of this study, after which
attention is shifted to a consideration of specific items of sector and
national wealth which require a decision as to inclusion in or exclusion
from the national balance sheet.
Once it has been decided what things are to appear on the balance
sheet, it is necessary to decide on what basis they shall be valued. This
Balance of payments figures are already integrated into national
income and product figures. Of the social accounting systems mentioned
earlier, only input-output remains unassimilated in this study. Given
the present unsettled state of input-output work and the virtual in-
fancy of the balance sheet, this situation is regrettable but probably
unavoidable. Fortunately, input-output has the least direct connection
with the balance sheet, though some incidental points of contact are men-
tioned in the course of the study.
question is the subject of Chapter II. Various possible methods of
valuation and the relationships between them are discussed. Also
treated is the question of divergences in asset valuation from the
viewpoints of individual economic units, sectors, and the nation.
Chapter III takes up the question of dividing the economy into
sectors for purposes of preparing the national balance sheet. The
necessity for analysis at the sector level rather than the national
level has already been suggested and is treated in more detail.
Various criteria for sectoring are examined, as are the types of
sectoring employed in other social accounting systems, especially the
Chapter IV presents sample balance sheets for the nation as of
the end of 1949 and the end of 1950. The balance sheets are prepared
on the basis of decisions as to scope, valuation, and sectoring,
which decisions are discussed in Chapters I, II, and III respectively.
Both years are included so that the relationships between income, pro-
duct, and funds flows for 1950, and the opening and closing balance
sheets for that year may be discussed in Chapter VI.
One of the conclusions of Chapters I and II is that the wealth
of individuals and of sectors can be satisfactorily related to the
national wealth through a process of cancelling claims of domestic
economic units against the corresponding liabilities of other units.
Some of the special problems arising in this process-which, following
accounting practice, we shall call consolidation, are discussed in
Chapter VI discusses the various relationships between flow
measures (national income and flow-of-funds) and stock measures (the
balance sheet) and, using 1950 figures, actually carries through a
reconciliation between the three types of statements for that year.
To the writer's knowledge this is the first time that such a process
has actually been completed using official and semi-official national
income and flow-of-funds figures. Chapter VII shifts to a more theo-
retical vein; some facets of economic and monetary theory are briefly
reexamined in the light of the relationships suggested by the balance
sheet and the manner in which it is linked to the more familiar
social accounting statements.
In summary, the writer believes that the preparation and regu-
lar publication of national balance sheets would add considerably to
our knowledge of economic processes within the nation. The study
which follows attempts to formulate the conceptual framework of the
balance sheet, to clarify many of the issues involved, and to propose
solutions for the problems which arise. The reader may judge how
well these objectives have been accomplished.
10A recent article by an accountant has proposed the prepara-
tion of a national balance sheet and independently arrived at essenti-
ally the same mechanics of consolidation as are here proposed. See
S. C. Yu, "National Position Statement: A Proposal on Operational
Principles and Process," Accounting Review, XXXIV (January, 1959), 74-
THE SCOPE OF NATIONAL AND SECTOR BALANCE SHEETS
A consideration of the scope of the national balance sheet--that
is, those things to be included or excluded--can lead one almost im-
mediately into some of the most difficult problems in the field of Eco-
Balance sheets deal with assets and liabilities, and they either
measure or have some close relationship to the notion of wealth. Like-
wise, they, in some sense, measure capital, regardless of which of the
numerous meanings of the term be accepted. Furthermore, they have a
definite relationship to measures of income--a concept notoriously en-
cumbered with ambiguities and varying definitions.
Without pretending to catalogue all of the difficulties which arise
in properly delineating national and sector wealth we may enumerate a
few of the more important ones. What, for example, is wealth? If
wealth can be defined for an individual, can it be defined in the same
way for a nation? If not, what are the relationships between the two
concepts? Is the term "asset" as used by accountants, identical with the
notion of wealth, and, if so, how do both terms relate to the idea of
A brief survey of the literature of economics from the last two
centuries is not conducive to optimism as to an ultimate resolution of
these issues. Attempting to define wealth, for example, is closely
related to the quest for the proper limits to the subject matter of eco-
nomics itself, an enterprise not marked with conspicuous success. Nor
can the whole matter be dismissed as an unfortunate succession of seman-
tic quibbles; there are some issues of substance involved.
Fortunately, what is required here is a somewhat more modest
achievement than the solution of apparently insoluble problems. Our
problem may be simply stated as follows: it is the practical one of de-
ciding which things belong on balance sheets for individuals and for the
nation and which things do not, within the framework of a system of
social accounting. Practical problems admit of practical solutions. We
are concerned with a social accounting statement for the United States
at a particular period of time, not with developing concepts of wealth
valid for all societies in all times--past, present, or future.
It is, of course, impossible to avoid completely all of the theo-
retical issues, no matter how "practical" a given solution. Indeed, the
proposals advanced here will conform closely to what most economists mean
by the term "wealth" and to what most accountants mean by the term
"assets." Moreover, we hope to demonstrate that a properly designed and
executed statement of individual and national wealth can show at the same
time various concepts of wealth or assets and the relationships between
them, which relationships are sometimes as significant as the items of
wealth themselves. The plan of attack, then, is to discuss some of the
general aspects of the measurement of balance sheet items and then to
move on quickly to a consideration of individual cases.
Wealth may be (and has been) defined in numerous ways from the very
broad to the very narrow.1 Some writers, speaking very broadly, have
considered wealth to consist of all desirable events or sources of desir-
able events. Such a notion extends far beyond the customary limits of
economics; it would include, for example, health, friends, memories,
literature, democracy, freedom--even civilization itself. Nor is this
nebulous collection narrowed down very much by stiffening the membership
requirements with the aid of that old economic standby--utility. All of
these things have utility.
One approach to a concept of wealth suitable for economic analysis
has been to include only tangible economic goods as wealth--economic goods
having already been defined in terms of relative scarcity and utility.
While possibly suitable for identifying most of the components of the
national wealth, this approach omits important segments of the wealth of
individual economic units or even large groups of units. Much of such
wealth is held in non-material form--that is, claims against other eco-
nomic units. Examples of such claims are bank deposits, mortgages, bonds,
and stocks. Even at the national level, a concept of wealth limited to
tangible goods cannot cope with intangible claims against foreigners or
claims held by foreigners against the nation.
Another attempted solution proceeds by dividing all of the com-
ponents of the very broad concept previously mentioned into categories
INo attempt has been made to cite references in the voluminous
literature for the ideas discussed in the following paragraphs. Most
of them are by now the common property of economists. For a susiary of
the development of various ideas of wealth through the early years of
the twentieth century see Irving Fisher, The Tature of Capital and
Income (New York: Macmillan Co., 1912).
of "transferable" (or "exchangeable") and "non-transferable" and then re-
garding only transferable items as comprising wealth for purposes of eco-
nomic inquiry. This procedure has at least the virtue of excluding human
beings (since they can no longer legally be bought and sold) and excludes
most of the other things which cannot be brought under the measuring rod
of money, things which economists generally prefer to leave outside the
scope of their attention. It may, however, exclude too much, especially
with respect to some things usually owned by governments. Government
buildings, for example, are not, as a practical matter, usually transfer-
able, yet they meet all of the other attributes one might require of
wealth--they are scarce, useful (usually), and tangible.
The relationship between the accounting term "assets" and the eco-
nomic term "wealth" touches upon some of these same questions and bears
closer examination. It should be pointed out immediately that the term
"asset" is susceptible to broad definition as is the case with wealth;
one hears it said that "Jones is an asset to the company," a statement
which may be literally true. Jones does not, however, appear on the
company's balance sheet; accountants have developed some fairly rigorous
conventions as to what things shall be included or excluded on business
In this connection, it may be observed that both assets and items
of wealth have a past, a present and a future dimension. They are "worth"
something now (at that particular point in time which constitutes the pre-
sent), they are the cumulative result of past efforts, and they give rise
to a flow of services in the future. These facets are clearly related,
since any item which will not give rise to useful services in the future
(however short the time period in the future) cannot be worth anything
now. Yet in another sense there is a vital difference, and the failure
to recognize this difference lies at the root of some of the difficulties
These three dimensions are recognized in most present day account-
ing by generally regarding assets as unexpired costs. The balance-sheet
value for the present is the sum of historical costs not yet yielded up
as a flow of services in the productive process. The flow of services
which will proceed in the future from a given item of wealth becomes for
accountants a future flow of past costs. Assets to the accountant are de-
ferred costs and nothing more.
It appears, then, that the accountants' notion of assets and the
economists' notion of wealth will coincide only under very restrictive
conditions not likely ever to be realized in practice. Most of the differ-
ences, however, exist with respect to valuation, and these are discussed
in detail in the succeeding chapter.
The relationship between the concepts of wealth and capital like-
wise has a somewhat confusing history. Some writers regard the two terms
as virtually synonymous. Others maintain a distinction between repro-
ducible goods (capital) and non-reproducible goods (land, or natural re-
sources). Another question is concerned with the productivity of the
goods in question or the lack thereof. If by capital is meant only goods
See, for example, Simon Kuznets, "On the Measurement of National
Wealth," Studies in Income and Wealth, Vol. II, by the Conference on Re-
search in Income and Wealth (New York: National Bureau of Economic Re-
such as industrial machinery actively engaged in producing other goods,
what of so-called consumer durable goods? If not capital, are they
wealth, or should they be excluded from wealth too?
The preceding paragraphs have, by design, raised some old issues
without yet attempting to resolve them. We may now begin to indicate some
of the solutions appropriate for a national balance sheet which is part
of a consistent system of social accounting. We shall first indicate some
of the guide lines which will be useful in attempting to specify the treat-
ment for particular issues, then move on to a discussion of some of these
First, let us develop a definition of wealth for an individual eco-
nomic unit. The initial step is to decide what shall be regarded as its
assets and then to decide on appropriate monetary values. Next, a de-
cision is required as to what shall be regarded as its liabilities and at
what values. For the difference between assets and liabilities, we shall
accept the accounting concept of net worth. Since it is an open question
whether liabilities may be regarded simply as negative assets, we shall
not adopt the tempting expedient of identifying net worth directly with
wealth, but, rather, we shall use the term wealth to refer to the whole
complex of assets, liabilities, and net worth. If pressed for a number
to signify a unit's wealth, we should give its net worth but always along
with the supplementary information afforded by its liabilities, hence, by
addition, the total of its assets as well.
For the nation we shall do essentially the same thing--decide on
national assets and their valuation, national liabilities and their valu-
ation, and regard the difference as national net worth. This entire com-
plex of national assets, liabilities, and net worth we shall designate as
the national wealth. In other words, we are proposing an accounting
approach, albeit a "social" kind of accounting. When national assets and
liabilities have been defined, then so, for our purposes, has national
A question arises immediately with respect to the relationship be-
tween the wealth of individual units and national wealth. Can national
wealth be regarded as the sum of the net worth figures of all individual
economic units? In order to answer this question it is necessary to make
a number of unoriginal but indispensable observations.
Economic units may bL classified generally as intermediate or ulti-
mate holders of wealth.3 Intermediaries hold assets, but they in turn are
owned by other economic units. Households, government units, and private
groups of individuals are the ultimate owners of a nation's wealth,
though much of their ownership is exercised indirectly--that is, through
the medium of claims.
Claims--of which bank deposits, mortgages, bonds, and stocks are ex-
amples--always appear at two points in the economy--at least those claims
which are domestically owned and owed. They appear as assets of some eco-
nomic units and as liabilities of others. In preparing a statement of
national wealth, domestic claims may be cancelled against the corresponding
liabilities, so that the only items remaining are assets which are not the
liabilities of other domestic units. In general, tangible assets and
3Goldsmith, "Measuring National Wealth in a System of Social
Accounting," loc. cit., p. 35.
claims against foreigners less foreign claims against domestic units are
the only wealth items which survive the process of cancellation.4
If we begin with balance sheets of all of the individual economic
units in the nation and, through the process of consolidation, work to-
ward the national balance sheet, any divergence between individual and
national wealth is limited to the following cases:
1. Assets or liabilities which appear on the balance sheet of
individual units but which do not disappear in the consoli-
dation process and which are not in reality assets or li-
bilities of the nation.
2. Bases of valuation which differ for the same asset or liability
as we move from the individual unit to the nation.
3. Claims which are valued differently as assets than as
Cases (2) and (3) above are considered in detail in Chapters II
and V; Chapter II deals with valuation problems, and Chapter V is con-
cerned with problems in consolidation. Case (1) appears as a possibility
with respect to certain types of assets such as patents, trademarks, copy-
rights, and goodwill and is discussed in more detail below.
An important consideration is consistency with the national income
and product measures. For an individual economic unit, income less current
expenditure (i.e., saving) results in an increased command over resources,
This process is frequently used in private accounting in preparing
group statements for a parent company and its subsidiaries and is known
as 'consolidation." This term will be used henceforth throughout this
or, from the standpoint of that unit, increased wealth. Similarly, we
should so organize the national income and expenditure statement and the
national balance sheet that national income less consumption (i.e.,
national saving) is equal to the increment in national wealth. That
part of current production which is set aside to augment stocks of
national assets (i.e., national investment) should likewise equal the
increase in the value of national assets, when allowance is somehow made
for changes, if any, in the values of existing assets. Some of the impli-
cations of these consistency requirements are discussed in Chapter VI
Another consideration which will influence the form and content of
the national balance sheet is the occasional desirability of showing con-
troversial items several ways, so that the results of alternative treat-
ment may be compared. If used too frequently, this device can make a
statement overly cumbersome; used sparingly, it presents valuable
Still another consideration is conformity to what individual eco-
nomic units ordinarily regard as constituting their wealth. This is not
an infallible guide since some units may value highly things which other
units ignore or regard as nuisances. Yet, if measures of wealth are to
have any motivational significance, we must not depart too far from notions
of wealth which, however vague, are held by the economic units in question.
Having established a working definition of wealth and enumerated
some of the considerations involved in identifying assets and liabilities,
we turn to a discussion of some specific issues. We are, in effect,
solving the problems of the scope of national and sector wealth in a
piecemeal fashion. Each decision regarding specific assets and liabil-
ities is a decision as to the scope of wealth in our social accounting
It may be well to begin at a point where virtual unanimity pre-
vails and then move on to more controversial areas. Accordingly, we may
say that nearly everyone regards machinery and equipment directly used
in the production of other goods (capital goods narrowly defined) as
wealth. Likewise, privately owned structures in cmonercial or industrial
use are universally treated as forms of wealth (and capital). Land is
also widely accepted as a form of wealth, whether treated as a species of
capital broadly defined or as a form of wealth in its own right. Like-
wise, stocks of goods (inventories) in the production and distribution pro-
cesses are usually regarded as both capital and wealth. All of the above,
needless to say, are also regarded as assets from the viewpoint of busi-
ness accounting, and additions to stocks of such goods are treated as in-
vestment in national income accounting. One cannot go much farther than
these cases, however, without getting into controversial territory. The
more important items which possibly qualify for inclusion in national and
sector balance sheets are discussed in the following paragraphs.
A problem of long standing in social accounting has been the treat-
ment of consumer durable goods such as automobiles, furniture, appliances,
and other relatively long-lived consumer articles. For purposes of
national income accounting, expenditures for such goods have usually been
regarded as consumption expenditures, not as saving or additions to the
stock of capital goods.5 If, on the other hand, balance sheets were pre-
pared for households, consumer durable goods would surely be listed as
assets. Household liabilities are frequently assumed in connection with
the acquisition of these goods, so that failure to list them as assets
results in a distorted view of household finances. Durables have "value";
they can be sold; in fact, they are considered sufficiently valuable to
secure the debts created when they are purchased. They meet the business
accounting criterion of deferred costs--their purchase takes place in one
time period, but the services which they produce are yielded up in later
time periods as well. If consumption of the services for which durables
are procured is measured solely by the initial purchase, consumption in
the period of purchase is overstated, and consumption in later periods is
The wealth or asset aspect of consumer durables, or the distortion
of consumption expenditure resulting from the present treatment of
national income, has been noted by Bandeen,6 Farrell,7 Fuerst,8
51t is interesting to note that the standard accounting system
recommended by the United Nations justifies this procedure by the lack of
the necessary statistical information. See United Nations, Department of
Economic Affairs Statistical Office, A System of National Accounts and
Supporting Tables (New York: United Nations, 1953), p. 9.
6Robert N. Bandeen, "Automobile Consumption, 1940-1950," Economet-
rica, XXV (October, 1957), 239-248.
7M. J. Farrell, "The Demand for Motor Cars in the United States,"
Journal of the Royal Statistical Society, CXVII, Part II (1954), 171-193.
8E. Fuerst, "An Alternative Presentation of Social Accounts,"
Accounting Research, III (July, 1952), 230-236.
Hamburger,9 Hough,10 Jacoby,11 Jones,12 Morgan,13 Stone,14 Derksen,15
Fabricant,16 Friend,17 Goldsmith,18 and Hicks.19 Other writers mention
the desirability of treating consumer durables as assets, and expendi-
tures for them as something other than consumption, but cite the lack of
data on useful lives and depreciation rates as the reason for continuing
the present national income treatment--namely, regarding purchases of
consumer durables as consumption expenditure. Among these writers are
9William Hamburger, "The Determination of Aggregate Consumption,"
Review of Economic Studies, XXII (1954-1955), 23-34. See also by the
same author "The Relation of Consumption to Wealth and the Wage Rate,"
Econometrica, XXIII (January, 1955), 1-17.
10Louis Hough, "An Asset Influence in the Labor Market," Journal
of Political Economy, LXIII (June, 1955), 202-215.
11Neil H. Jacoby, "The Demand for Funds by American Business Enter-
prises," Journal of Finance, III (October, 1948), 27-38.
12Homer Jones, "Some Aspects of Demand for Consumer Durable Goods,"
Journal of Finance, IX (May, 1954), 93-110.
13James N. Morgan, "The Structure of Aggregate Personal Saving,"
Journal of Political Economy, LIX (December, 1951), 528-534.
1R. Stone and D. A. Rowe, "The Market Demand for Durable Goods,"
Econometrica, XXV (October, 1957), 423-443.
15J. B. D. Derksen, A System of National Book-Keeping, National
Institute of Economic and Social Research Occasional Papers, X
(Cambridge, England: Cambridge University Press, 1946).
16Solomon Fabricant, Capital Consumption and Adjustment (New York:
National Bureau of Economic Research, 1938), p. 139.
17Irwin Friend, Individuals' Saving: Volume and Composition (New
York: Wiley, 1954), p. 14.
1Gold~sith, A Study of Saving, I, 28.
19J. R. Hicks, Value and Capital (2d ed.; Oxford: Clarendon Press,
1946), p. 176.
Edey and Peacock,20 Stone,21 and Vandermeulen.22
If we are to have a national balance sheet and a national income
and expenditure statement prepared on a consistent basis, the treatment
of consumer durables will have to be resolved in one direction or the
other. Either they are assets--hence, (for our purposes) wealth--so that
expenditures to acquire them are not current consumption; or they are not
assets at all, so that their purchases may be regarded as current con-
sumption. It will not do to continue the present national income treat-
ment of writing off such goods as current consumption and then resurrect-
ing them to a place on the national balance sheet.
Some of the difficulty stems, no doubt, from the national income
treatment of all private expenditures as either consumption or invest-
ment--investment being defined as additions to the stock of capital goods.
There is a reluctance, understandably enough, to call consumer durables
capital goods. If they are not capital goods, expenditures for them may
not be considered investment spending, hence must be consumption. This
dilemma is illustrated by the present national income treatment of
residential construction--such spending is clearly not consumption; yet
20Harold C. Edey and Alan T. Peacock, National Income and Social
Accounting (London: Hutchinson's University Library, 1954), p. 52.
21R. Stone, "Functions and Criteria of a System of Social
Accounting," Income and Wealth, Series I, by the International Associa-
tion for Research in Income and Wealth (Cambridge, England: Bowes and
Bowes, 1951), p. 17.
22D. C. Vandermeulen and A. J. Vandermeulen, National Inc'ime:
Analysis by Sector Accounts (Englewood Cliffs, New Jersey: Prentice-
Hall, Inc., 1956), p.11.
houses are not capital goods, defined in a sufficiently narrow way.
It seems, then, that most of the difficulties revolve around the
definition of capital. If the term is restricted to those produced goods
which are used to produce other tangible goods, then all sorts of goods
which produce only services, such as structures and equipment used in
transportation, wholesaling, retailing, entertainment--in fact, much of
our economic activity--are excluded. Surely this concept is too narrow;
expenditures for these things cannot be regarded as consumption expendi-
ture. Yet, if we admit goods which produce only services to the status
of capital, we are obliged to admit consumer durables, for they too pro-
We might restrict capital to include only those produced goods
which produce goods or services to be sold in markets. This is not a
wholly satisfactory solution either, since by this criterion, houses
which are rented to others are capital, yet identical houses which are
occupied by their owners are not. Stoves and refrigerators in restaurants
or furnished by landlords to tenants are capital, but those used by their
consumer-owners are not. Chevrolet taxicabs are capital, but Chevrolets
in general are not. Automatic washing machines in laundromats are capi-
tal, while identical machines at home are not. These inconsistencies
suggest the need for another definition of capital goods.
Another possibility is to be guided by durability--those goods
which give up their services over sufficiently long periods of time may
be regarded as capital goods, while those which are quickly consumed are
not. This criterion does not, however, provide any grounds for excluding
consumer durables, since some have useful lives longer than much indus-
Other criteria could be tried, perhaps, and found wanting, but the
futility of such a pursuit should now be apparent. There is probably no
single definition of capital which is wholly satisfactory; at least the
endless (and somewhat inconclusive) literature on the subject suggests as
Perhaps the best solution for our purposes is to recognize two kinds
of capital, which we might call "business capital" and "consumer capital."
The latter would include consumer durable goods and residential structures,
while the former would coincide with the present national income defini-
tion of capital except for residences. There would correspondingly be
two kinds of investment which we might call "business investment" and
"consumer investment."23 Some problems remain, of course, and there are
no universally applicable rules. We shall have to be, in some cases,
purely arbitrary, but there is in social accounting ample precedent for
arbitrary choice. For example, are fifteen-story apartment houses con-
sumer or business capital? If they are treated as consumer capital,
what happens if the first floor is used for retail outlets? These prob-
lems are somewhat academic at the moment, however, since estimates of
national assets which may become available in the foreseeable future are
not likely to be sufficiently refined so that such difficulties will
James N. Morgan, "Consumer Investment Expenditures," American
Economic Review, XLVIII (December, 1958), 874-902. The above approach
is suggested by this recent article.
Another issue which has been a source of continuing controversy
is the treatment of durable goods purchased by government units. This
is not a particularly serious problem for government enterprises which
sell their services to the public and which have counterparts in private
business. The difficult questions arise with respect to highways,
roads, streets, sidewalks, sewer systems, bridges, harbors, parks, air-
ports, and public buildings. These items are presently treated as
current (i.e., non-capital expenditures) in national income, along with
all other expenditures of governments, as, for that matter, are capital
expenditures of public enterprises. They meet most of the usual criteria
for capital goods--they are durable, and they produce useful services.
Their services are not generally sold in markets, however, though excep-
tions such as turnpikes and airports come quickly to mind. As we have
already seen, this last consideration is not conclusive anyway in de-
termining whether or not goods can be classified as capital.
The wealth aspect of public property such as that described above,
or the desirability of showing government capital expenditures separate-
ly in the national accounts has been noted by Dray,24 Fuerst,25 Goode,26
F. Sewell Bray and Richard Stone, "The Presentation of the Cen-
tral Government Accounts," Accounting Research, I (November, 1948),
25Fuerst, "An Alternative Presentation of Social Accounts,"
26Richard Goode and E. A. Birnbaum, "Government Capital Budgets,"
International Monetary Fund Staff Papers, V (February, 1956), 23-46.
Jacoby,27 Derksen,28 Marshall,29 Fabricant,30 Kuznets,31 and
Goldsmith.32 A staff member of the National Income Division of the Com-
merce Department acknowledged the desirability of full-scale capital
accounting for government assets--that is, treating such expenditures as
gross investment, computing capital consumption, and deriving net invest-
It might be well to recognize that such investment is different in
some important respects from business investment, especially as to motiva-
tion and cyclical behavior. We shall, accordingly, designate durable
goods owned by government units as "government capital" and the expendi-
tures involved in its acquisition "government investment."
Durable military goods pose a further problem. To be consistent,
we should probably regard them as assets, since they are durable and pro-
27Jacoby, "The Demand for Funds by American Business Enterprises,"
28Derksen, op. cit.
29Alfred Marshall, Principles of Economics (8th ed.; New York:
Macmillan Co., 1952), p. 59.
30Fabricant, op. cit., p. 120.
31Simon Kuznets, "Government Product and National Income," Income
and Wealth, Series I, by the International Association for Research in
Income and Wealth (Cambridge, England: Bowes and Bowes, 1951), p. 199.
32Goldsmith, A Study of Saving, I, 113.
33George Jaszi, "The Conceptual Basis of the Accounts--a Re-examin-
ation," A Critique of the United States Income and Product Accounts
(Studies in Income and Wealth, Vol. XXII), by the Conference on Research
in Income and Wealth (Princeton: Princeton University Press, 1958), p. 78.
vide services applicable to several time periods; quantitatively they
can hardly be ignored, at least in recent decades. The real problem
arises in measuring the length of time over which they provide useful
services. In wartime--when, presu-mably, their usefulness is at a maxi-
mam--their useful lives are most uncertain; in peacetime they are sub-
ject to rapid obsolescence, and their opportunity costs, once produced,
are virtually zero. There would appear to be little to gain by going
through the motions of wealth accounting for such items; at best, they
should probably be shown as supplementary information.
Monetary gold is treated in the national income accounts as part of
net foreign investment--that is, a claim on foreigners. It should be
treated as a domestic tangible asset for balance sheet purposes; not an
international intangible asset; foreign nations recognize no liability to
the United States in the amount of our gold holdings as they would for
actual claims against them. Monetary silver may properly be regarded as
an asset, though the appropriate valuation is another question which is
deferred to Chapter II.
The point has now been reached where most of the major components
of national wealth have been discussed; further treatment of less im-
portant items runs into rapidly diminishing returns. In the interests of
completeness, however, a few such items will be mentioned.
For the record, stocks of perishable and semi-durable goods in
households--such as food, clothing, and miscellaneous supplies--should
probably be included in national wealth. Moreover, they are quantitative-
ly more important than one might think.34 Year-to-year changes are
probably small, however, so that for purposes of analysis it makes little
difference whether or not such goods are included.
Such items as works of art and historical monuments present interest-
ing conundrums; as a practical matter it makes little difference whether
they are included or not. Any income to which they give rise is rarely
measurable in monetary terms so that they might well be omitted from
monetary measures of national wealth. Perhaps art treasures for which a
market exists should be included since a reasonably appropriate measure
of money value exists.
With respect to claims against foreigners and claims owed to
foreigners, the principal problem for purposes of determining the scope
of the national wealth is to identify foreigners. The best practice
seems to be to follow the conventions developed by the National Income
Division and followed by the Flow of Funds Section. Territories and
possessions of the United States are excluded from the national accounts
(hence treated as foreigners) as are the transactions of United States
citizens residing permanently abroad. Residents of the United States
proper are included, whether citizens or not,
Technically speaking, if only assets and liabilities of domestic
economic units appear on the national balance sheet, physical assets lo-
3Goldsmith quotes a study by Lenore A. Epstein which indicates
holdings of semi-durables such as shoes and clothing of $34 billion in
1946. See Raymond W. Goldsmith, "A Perpetual Inventory of National
Wealth," Studies in Income and Wealth, Vol. XIV, by the Conference on
Research in Income and Wealth (New York: National Bureau of Economic Re-
search, 1951), p. 36.
cated in the nation but owned directly by foreigners would not appear at
all. In practice, it would be better to show such assets on the nation-
al balance sheet, with an offsetting liability to foreigners in the same
amount so that all domestically situated wealth is reported. Most
foreign ownership, however, is in the form of claims against domestic
economic units, including shares of domestic corporate subsidiaries of
foreign firms, so that this problem does not arise. Certain other prob-
lems arise in connection with foreign assets and liabilities, but they
are concerned with the classification of assets and liabilities (dis-
cussed in Chapter IV) and valuation (discussed in Chapter II).
The foregoing discussion has treated all of the items which are im-
portant from the viewpoint of the nation. We turn now to a discussion of
some of the items which become significant only for individual economic
units or for sectors. Sector balance sheets will contain many items that
do not appear on the national balance sheet; specifically, all of the off-
setting claims and liabilities already mentioned. Most of these items
present no problems as to inclusion or exclusion; the troublesome
features appear with respect to uniform valuation and are discussed in
Chapters II and V. There are, however, a few matters which warrant sep-
arate attention; these are presented below.
Certain intangible business assets such as patents, copyrights,
trademarks, and goodwill appear on business balance sheets. They will
not disappear in the process of consolidation since no one else in the
economy regards them as liabilities. Can they properly be entered, then,
on balance sheets for sectors or for the nation? From the standpoint of
business firms, some of these items constitute saleable wealth, though
rarely at the values which appear on balance sheets, where they may be
listed at nominal values or omitted entirely. Part of the problem is one
of valuation and as such will be discussed in Chapter II; the present con-
cern is whether they should be included or not.
From the standpoint of the nation, useful knowledge of the sort
represented by patents is an enormously valuable asset. There is, how-
ever, no way of valuing the total of knowledge available to a nation, so
that showing on the national balance sheet that tiny portion of knowledge
which happens at the moment to be covered by outstanding patents is a
futile and meaningless gesture. Moreover, if patents are considered as
items of value to firms, it is because there are associated with them ele-
ments of monopoly earning power. There could, theoretically, be shown on
the balance sheets of the prospective purchasers of the patented product
a liability which measures the present value of future "excessive" prices
to be paid--prices which exceed those which would prevail under competi-
tive conditions. This "excess" is, of course, the thing which gives rise
to the asset value of the patent in the first place. If such liabilities
were entered, they would cancel against assets in the consolidation pro-
cess so that no asset value would remain for the national balance sheet.35
In any event, patents should not appear on the national balance sheet as
part of the national wealth, a conclusion which may be reached by either
3This point is suggested by Goldsmith's discussion of the same
problem. See his "Measuring National Wealth in a System of Social
Accounting," loc. cit., p. 46.
of the above lines of reasoning.
From the viewpoint of individual economic units, however, patents
are elements of wealth. They have value and can be sold. They should
be permitted to appear on the balance sheets of the owning units, since
their omission would distort the actual economic situation of the patent
Much of the above discussion of patents applies to the other in-
tangibles mentioned. In general, if the item represents a marketable
possession--that is, if there is actually an opportunity cost involved
in not selling--it should appear as an asset in individual and sector
balance sheets but not in the national balance sheet. There are, of
course, some very real problems of valuation, but these, as noted pre-
viously, are treated in the succeeding chapters.
With respect to households as a sector, some problems appear in
connection with pension plans. Should the present values of future bene-
fits, however computed, appear as household assets? In the national in-
come scheme of things, different answers are given depending upon
whether the pensions in question are publicly or privately administered.
Contributions by employers to private (self-administered and insur-
ance company) pension funds are included in personal income (as "other
labor income" under salaries and wages). Since neither these nor
individuals' contributions appear as consumption expenditures, these
amounts are implicitly reflected in personal saving, hence as increases
in household assets. Furthermore, earnings of these funds are included
in personal income (explicitly for self-administered plans and by imputa-
tion for insurance companies) so that cumulated amounts in the funds
have all appeared in personal saving at one time or another.
It is true that accumulated funds, which appear as assets of the
insurance sector in this study (see Chapter III) and also as liabilities
to households, would not match the assets on household balance sheets if
the latter were computed actuarially, mostly because past service credits
have not yet been fully funded. With this exception, however, household
assets include the present values of future benefits, which also appear
as liabilities of the insurance sector. For purposes of this study,
accruing claims for private pension benefits will be regarded as house-
hold assets and liabilities of the insurance sector.
A similar problem exists with respect to the social security pro-
gram. Households are receiving or will in the future receive old age,
survivors', and disability benefits. Those benefits already "earned" by
covered employment to date might be computed, and the present value of
this amount discounted at a suitable interest rate might be entered as
household assets, and liabilities of the social security trust funds.36
Another chain of reasoning which leads to quite different results
is to consider the benefits as just another government transfer payment
for which no specific liability is accruing any more than a liability
for future federal payrolls is accruing. That the benefits happened to
be financed in a particular way is incidental; social insurance contribu-
36Goldsmith's savings study regards ct-rulated holdings of the
various pension and retirement funds as household assets, and annual
additions as a component of household saving. See A Study of Saving,
tions could be regarded as just another form of taxation. Households
would, under these assumptions, show no asset for future benefits, nor
would the government trust funds show any such liability. This is the
treatment suggested by the national income accounts, which do not show
contributions to the social security trust funds as personal saving or
disbursements as personal dis-saving.
It is highly improbable that any households consider the present
value of future benefits likely to be received as an asset--at least
not an asset on the order of a savings account or savings bonds. Such
an "asset" is highly illiquid; it is strictly non-transferable. It is
possible, on the other hand, that the likelihood of benefits in the
future has some bearing on present saving and spending decisions, though
how much, if any, is wholly a matter of conjecture.
The national wealth total is not, of course, affected by the treat-
ment of this problem. Household assets and government liabilities are
affected, however, and by a very considerable sum. In view of the fact
that the liability of the trust funds is not and probably never will be
fully funded and that the assets involved for households are at best
nebulous, the national income treatment will be followed in this study.
For those who prefer the alternative treatment, one figure--namely, the
estimated government liability to households--can be shown as supple-
mentary information and will suffice to make the necessary adjustment.
Another problem worthy of comment is the treatment of tax liabil-
ities--both those which have already accrued and those which will fall due
in the future. With respect to those tax liabilities already payable,
they should, in the interest of completeness, be shown as assets of
governments and liabilities of other economic units. The national in-
ccme accounts already follow this procedure insofar as corporate profits
taxes are concerned in order to maintain consistency between corporate
income statements and the national accounts. The implicit result is
that changes in accrued taxes receivable by governments are an element
of government saving or dis-saving. Furthermore, important changes such
as the recent shift of corporate taxes toward a "pay-as-you-go" basis
are obscured if a cash rather than accrual basis is followed.
With respect to future tax payments, it might be argued that tax-
payers' balance sheets should show a liability amounting to the present
value of such future payments. This argument is sometimes advanced in
connection with the national debt; if it is ever paid off, taxes will
provide the necessary funds. The national debt is a liability of tax-
payers; their balance sheets should recognize it as such.
This argument may be countered at several levels. By setting up
accounts for governments as separate economic units we recognize that
some assets are collectively owned and some liabilities (such as the
national debt) are collectively owed. Balance sheets of government units
are the appropriate places for such assets and liabilities to be shown.
Going beyond this to taxpayers' balance sheets is not necessary since it
is recognized by everyone that government assets and liabilities are
shared by all the people, nor is such a procedure feasible since each tax-
payer's share of public debt is not determinable anyway.
Moreover, if we show future tax payments as present liabilities of
taxpayers, we must show as assets the present values of future incomes
from which the taxes will be paid. For labor income this amounts to
showing as an asset the present capitalized value of future earnings,
which we have already excluded from the national balance sheet. Societies
differ as to what they regard as wealth, but in Western societies, at
least since the abolition of slavery, human beings are generally ex-
cluded from any notion of measurable wealth as assets.
With respect to property incomes, balance sheet values of earning
assets generally tend toward capitalized present values of future earn-
ing; to the extent that taxes reduce earnings, present balance sheet
values already reflect an allowance for future iax payments. Regardless
of the source of income from which taxes are to be paid, then, it would
be inappropriate to record any liabilities for taxpayers.
We have completed a survey of the more important types of assets
and liabilities that might appear on national and sector balance sheets.
We turn now in the following chapter to a consideration of the problems
involved in assigning appropriate values to these assets and liabilities.
VALUING SECTOR AND NATIONAL WEALTH
The question of placing a proper valuation on objects of wealth
has plagued economists and accountants (and others) for years. Had it
been satisfactorily resolved, the literature of both disciplines would
be a great deal slimmer, for there are innumerable related problems
which would be solved as well. This chapter, then, does not pretend to
give easy answers in a few pages to such perplexing questions; even a
mere summary of the present state of professional opinion would consti-
tute an undertaking as large as this entire study and is, of course, not
attempted. Some discussion of valuation, however, is inevitable for
developing a national balance sheet; the aim of this chapter is to
approach the subject of valuation from the standpoint of a social
accounting measure of sector and national wealth.
In our discussion of valuation problems it will be helpful to deal
individually with different types of assets and liabilities. We shall
divide wealth into three familiar categories--namely, reproducible
tangible assets, nonreproducible tangible assets, and intangibles. Each
is discussed in turn.
First treated is the category of reproducible tangible assets, which
is in economics usually discussed under the heading "capital goods" and
in accounting as plant (excluding land), equipment, and inventories. In
many respects, valuation problems are most acute with respect to these
items, and some of the principles developed will apply to other cate-
gories of wealth as well.
The usual alternative bases of valuation for reproducible tangible
assets are historical cost, reproduction cost, replacement cost, and
market value. Replacement cost and reproduction cost are technically
different, though they are frequently confused in the literature. Re-
production cost suggests the cost of producing another article exactly
like the one in question, whereas replacement cost suggests the cost of
acquiring a good which will perform the same service. This distinction
becomes important during periods of rapid technological advance--which
is to say nearly all the time under present conditions.
The writer admits in advance to a predilection for market value as
the basis of valuation appropriate for use in the national balance sheet.
Unfortunately, there are a number of compelling reasons, discussed in de-
tail below, why this method is not feasible in practice. Accordingly,
depreciated original cost adjusted to depreciated current cost by means
of appropriate indexes is proposed as the best approximation to market
value. As a practical matter, the result is strictly speaking neither
reproduction cost nor replacement cost, since price indexes too have
trouble with technological change. It is rather a mixture of the two
versions of cost, which is in the writer's opinion sufficiently precise
for the purposes of the national balance sheet.
Since the proposed method of valuation for the national balance
sheet differs from that used in private business accounting, some con-
siderable discussion is devoted to this matter of alternative valuation
bases. Business accounting provides much of the basic data for social
accounting; for that reason it is necessary to justify in some detail
any departure from the carefully developed body of generally accepted
It will be instructive at the outset to consider a set of highly
restrictive conditions under which all of these various bases of valua-
tion produce the sane results. Let us operate for a moment at that
rarefied level of abstraction represented by a perfectly competitive
economy in a state of long-run equilibrium--no net investment is taking
place, nor does technology advance, and population is constant. Perfect
foresight exists, so that divergences between planned and realized mag-
nitudes are not possible, and there is no such thing as risk, so that
debt and equity funds yield identical returns. Moreover, the interest
rate, identical for all capital goods, is determined by the marginal pro-
ductivity (or marginal efficiency) of capital since all of the motives
for holding cash disappear with risk and uncertainty except the trans-
actions motive. With no liquidity preference as such, there is no
liquidity preference theory of interest.
In this highly imaginary economy, consider a certain capital good
which costs $100.00 and yields a product whose sales, after deducting
the costs of the other factors of production, yield a return of exactly
$23.74 per year for five years. Assume that the machine is purchased and
installed on January 1, year 1, and that the product is completed and sold
once a year for each of five years--December 31 of years 1, 2, 3, 4, and
5. Each December 31.after the product is sold, the books are closed,
depreciation is recorded, and profit is computed.
A glance at an annuity table will indicate that the marginal
efficiency of this piece of capital is 6 per cent--that is, the rate re-
quired to discount the returns to the cost of the machine is 6 per cent.
The relevant transactions may be simply represented in the following
RETURNS, UrRECOVERED COST, AND DEPRECIATION OF A
Unrecovered Gross Return
Before After Return as
Decem- Current Current Deprec- Net Per Cent of
ber Year's Year's iation Return Total Unrecovered
31 Deprec- Deprec- Cost
Year nation iation
1 $100.00 $82.26 $17.74 $6.00 $23.74 6
2 82.26 63.46 18.80 4.94 23.74 6
3 63.46 43.43 19.93 3.81 23.74 6
4 43.53 22.40 21.13 2.61 23.74 6
5 22.40 -0- 22.40 1.34 23.74 6
The purchaser of the machine, under the extreme conditions of no
risk and no uncertainty, is, in effect, buying an annuity which yields
the going long-period interest rate.* We are sacrificing some precision
iThere are some people who think the interest rate under long-run
competitive equilibrium must be zero. This is true only if the supply
curve of new savings happens to become zero only at a zero interest rate.
It would be just as plausible for such a rate to be positive or negative.
by letting the net return ($6.00 in Year 1) be compared to the year-end
value before depreciation ($100.00 in Year 1) to compute the yield of 6
per cent. This assumes that the year's depreciation happens all at once
late in the afternoon on December 31. We should more properly show de-
preciation occurring continuously throuSh the year and compute the net re-
turn on the average value of the asset during the year. Refinements
such as this, however, would not alter any conclusions of the argument
and would distract from the simplicity of the example.
Several observations are now in order. If, on January 1 of Year 2
one were asked what he would pay for the machine under the assumed condi-
tions, he would give for his answer the book value after the first year's
depreciation was deducted, i.e., $82.26. In other words, 6 per cent is
the rate which correctly discounts four payments of $23.74 for the next
four years to a present value of $82.26. Book value and market value co-
incide. Long-period equilibrium further implies no cost changes, so that
reproduction cost is equal to historical cost new, and depreciated repro-
duction cost would be equal to depreciated book cost. Moreover, the ab-
sence of technological change implies that replacement cost is the same
as reproduction cost. All of the bases of valuation, then, coincide
under these extreme assumptions.
It should also be noted that depreciation charges increase as time
passes; this result is implicit in the method adopted for computing the
We are assuming it here to be 6 per cent--that is, if the rate should fall
below 6 per cent, people will dis-save; hence shrink the supply of cap-
ital; hence raise its return.
marginal efficiency of capital and the assumption of constant gross re-
turns for each of the five years. Straight-line depreciation would, in
this case, carry the implication of a zero rate of interest. This
should not, however, be interpreted as a criticism of present-day
accounting practices which rely heavily on straight-line depreciation.
Our example is a long way away from a real world.
It is possible to arrive at the same conclusions even in the ab-
sence of long-run equilibrium, if the remaining assumptions--i.e., per-
fect competition, lack of technological change, constant population, and
no risk or uncertainty--are retained. The conditions of a dynamic equilib-
rium, as an economy under the above four assumptions moves toward long-
run equilibrium, have not been fully and satisfactorily worked out, but
they would very likely include the following:
1. Net investment is taking place; capital goods production
measured in current prices exceeds depreciation allowances.
2. There is a demand for capital funds on the part of business
firms, and a supply of same on the part of savers. Or, putting
the same thing in different words, there is a supply of se-
curities (stocks or bonds indifferently under our conditions
as noted above) by businesses and a demand for same by savers.
3. There is a corresponding demand by businesses for capital
goods and a supply governed by marginal costs in the capital
4. Returns to capital are identical in the consumer-goods and
capital-goods industries. That is, the demand for capital
goods to expand the stock of capital has been met by a
sufficiently expanded capital-goods industry so that ex-
cessively high returns do not prevail there.
5. The marginal productivity of capital--hence the rate of
interest--declines (though in a perfectly foreseen fashion),
but values of existing capital goods are not affected. This
is true because even though new capital goods through di-
minishing returns are forcing down the return to all capital--
new and old--the falling interest rate means that lower quasi-
rents are being capitalized at a correspondingly lower rate;
hence capital values need not change.2 The equality between
the various bases of valuation developed above need not be
It is not possible to drop any more assumptions, however, without
destroying this neat equality. In other words, if these highly restrict-
ive assumptions are not fulfilled, the different methods of valuation
lead to different results. We shall catalogue some of the departures
from the extremely restrictive assumptions given above and note the
effects on the relationships between original cost, replacement cost,
reproduction cost, and market value.
One such is a departure from competition either in the markets
2This would not be strictly true if all earnings were capital-
ized into present values. Falling returns to capital imply higher wages,
so that present values of human labor rise on two counts--higher returns
and a lower rate of capitalization. This observation is made by A. P.
Lerner in his article "On the Marginal Product of Capital and the Harginal
Efficiency of Investment," Journal of Political Economy, LXI (February,
for capital goods themselves or in the markets for goods produced with
their aid. A related market imperfection is the lack of widespread
knowledge of alternative opportunities as well as an infrequent market
turnover for most existing capital goods. The multitudes of omnis-
cient little competitors scurrying around taking advantage of every
minute departure from equilibrium exist only in the textbooks. Condi-
tions such as these explain why market value is not an overly attract-
ive basis of valuation for either the business accountant or the social
A changing general price level, assumed out of existence in long-
run equilibrium and not necessarily a feature of the dynamic equilibri-
um discussed above, in practice represents a frequent source of diver-
gence between the different valuation methods, especially between histor-
ical cost and the other three. Since the net movement of prices over the
last hundred years or so (at least) has been up, the usual result is
that historical cost is the lowest value of the four, though there are
important exceptions (such as the 1930's).
The existence of risk and uncertainty introduces a divergence be-
tween plans or expectations and realizations. Market value, which al-
ways depends on someone's expectations about future returns, is strictly
an ex ante concept. Cost (whichever version--historical, replacement,
or reproduction), on the other hand, is a realized, or ex post concept.
True, from the standpoint of the purchaser of a capital good, cost and
market value are identical at the moment of purchase, but this equality
never necessarily holds again.
Changes in technology or changes in consumer tastes are equally
destructive of the values of existing capital goods, a situation des-
cribed by accountants as obsolescence. The inevitable result of tech-
nological advance is to render market values (or in the absence of an
objective market value, present values reckoned by discounting expected
returns within a firm) lower than costs, except to the extent that ob-
solescence is anticipated and allowed for in depreciation rates. As
we have already noted, this factor (technological advance) also destroys
the equality between replacement cost and reproduction cost. Shifts in
consumer tastes raise market values of capital goods above costs or
force them below costs, depending on the direction of the shift. About
the only certainty is that some change will occur.
This brief recitation of departures from the abstract world of
perfect competition and equilibrium--all of which departures are persist-
ing features of our own and other economies--suggests that the four pro-
posed bases of valuation will normally lead to four different results
even if they all could be computed, and they frequently cannot. Further-
more, no one is intrinsically superior to the others, so that the reams
of paper and gallons of ink consumed in "proving" that one or the other
of the concepts is the "best" have been largely wasted. If one is inter-
ested in computing realized returns, he will use one of the cost con-
cepts; which one he uses will depend on what is to be measured as capital
gain or loss and what is to be considered returns from operations.3 If,
3The current controversy in accounting over historical cost versus
one of the other cost variants may be viewed as an argument about how
on the other hand, one is interested in motivational factors, he is mov-
ing into e; ante territory; market value, which best measures opportunity
cost, is the relevant consideration.
As noted in Chapter I, assets or items of wealth have past, pre-
sent, and future dimensions; a balance sheet struck at a given moment--
the "present"--is a link between past and future. If cost is taken as
the measure of past efforts embodied, and a stream of returns to be
earned represents the future dimension, then the value which appears on
the balance sheet is the present link. We have already seen, however,
that these differing dimensions converge on identical values only in
special circumstances never likely to be realized. Some sort of choice
must be made, and accountants have generally chosen the past.
Any such choice results in a dilemma; if a value which reflects
future expected returns is selected (i.e., market value), the balance
sheet gives no information about embodied costs. If, on the other hand,
unrt-overed costs are used as the basis, the balance sheet tells nothing
about the present worth or future expectations, except that when market
values fall below costs this fact is generally reflected in present-day
Unfortunately, non-accountants persist in the belief that balance
money gains shall be allocated between capital gains on the one hand
and profits from operations on the other. Historical cost devotees
want to call all realized returns profit, while advocates of, say, re-
placement cost as the proper valuation basis want to show some of the
returns as capital gains. The latter group, if consistent, will also
write up the asset values as well as increase depreciation charges over
those indicated by historical costs. This write-up is a capital gain
which is realized" through depreciation charges.
sheets (or statements of financial position, more precisely) tell
something about what an enterprise is "worth," a not wholly unreason-
able expectation. In response to this attitude, some accountants deny
that balance sheets do any such thing. Some have ventured to suggest
that since balance sheets are such technical affairs they should probably
not be entrusted to the public (which is bound to misinterpret them) in
their present form.5 The balance sheet becomes a mere appendage of the
income statement--a temporary resting place for costs which await their
final fate in some future income statement. This is a far cry from
former times when income statements served only to link together success-
ive balance sheets, which were thought to contain the really valuable in-
The shift of emphasis to the income statement is no doubt a
healthy thing; we have witnessed much the same development in national
accounting. As we have already pointed out, national income and product
figures deserve all the care and attention they get and then some. We
should not subordinate them to any of the other social accounting state-
Nor is there anything wrong with looking upon assets as deferred
costs. In the simplified example given at the beginning of this chapter,
See, for example, John N. Myer, "Fallacies in the Balance Sheet
Approach," Accounting Review, XXI (January, 1946), 8-13. In this
article accounting teachers are taken to task for spreading such false
5George 0. May, "The Future of the Balance Sheet," Journal of
Accountancy, LXXXVIII (August, 1947), 98-101.
assets viewed as deferred costs and assets viewed as present values co-
incided. When different methods of valuation diverge (as they do in
reality), especially when they diverge widely as in a period of rapidly
rising prices, to hold that assets are only deferred historical costs
is patently absurd. They are also the present value of a stream of ex-
pected future returns, and, if one has to choose which of the two con-
cepts is the more significant, he would choose the latter. The past is,
after all, past; the only decisions which can now be made refer to the
present and future.
There are, of course, good reasons for maintaining historical cost
as the basis for business accounting. Here, at least, is an objective
and verifiable standard and one which affords a real measure of protect-
ion to the public and business shareholders after the shabby and fraud-
ulent tricks of an earlier day. Most of its defenses are on practical
grounds, however; there are few good theoretical ones.
We may now return to the problem of valuing reproducible tangible
assets in the national balance sheet. Given the necessary information,
it would seem preferable to use market value as the appropriate basis of
valuation. Such a basis fairly represents the opportunity cost of hold-
ing and using assets, which is the relevant consideration as far as any
decisions in the present are concerned. It would state the national
wealth on the same basis as the national income and product. It would
provide a basis for calculating capital consumption allowances on a
current cost basis, a valuation basis more nearly comparable to other
gross national product components--a sorely needed improvement in the
national accounts. Capital formation figures, which may now sensibly
be expressed as a percentage of gross or net national product, could also
be expressed as a percentage of existing capital since both would be on
a comparable valuation basis. Moreover, stating each sector's wealth in
current market values allows an estimate of its current command over
Unfortunately, there are some formidable difficulties in using
market value as a basis of valuation, and they do not have much to do
with difficulties of gathering existing data, which are formidable in
their own right. Only a very small fraction of the nation's wealth
passes through markets in a given period of time, and much of the nation's
wealth never does at all. There are no such things as objectively deter-
mined market values in most cases.6
Our problem, then, is to use that basis of valuation which most
nearly approximates current market value. In the writer's opinion, the
appropriate second-best measure is original cost adjusted to a current
basis by means of specific indexes--that is, separate indexes for each
class of capital goods--with depreciation reserves similarly adjusted
by means of price indexes. Since the available indexes are based on
current production costs of similar articles, our method results in a
mixture of replacement and reproduction costs, depending on the extent
to which quality changes have been reflected in the indexes and the extent
6This difficulty is stressed by Kuznets, among others. See "On
the Measurement of National Wealth," loc. cit., pp. 24-27.
to which quality changes have affected values of existing capital goods.
The method outlined above, namely, original cost adjusted to current
cost by means of specific price indexes and depreciated on a current
cost basis, is used in this study.
This is not a wholly satisfactory solution. In fact, there is no
way, really, to measure the stock of capital; we are merely using the
best approximation. It follows that the same thing is true of capital
consumption allowances.7 The method outlined has been used by the De-
partment of Comnerce,8 and it is followed by Goldsmith in all of his
recent work.9 It is further suggested as an appropriate method in sev-
eral other recent articles.10
With respect to tangible nonreproducible assets--the most impor-
tant of which are land and other natural resources--the theoretical
problem is much simpler. Since there is, strictly speaking, no produc-
tion cost involved, the market provides the only means of valuation.
As a practical matter, the problem is more serious since only a small
7Keynes is one of the several writers who has pointed this out.
See John Maynard Keynes, The General Theory of Employment, Interest and
Money (New York: Harcourt, Brace & Co., 1936), p. 39.
8Though not, of course, in official income and product figures.
See Raymond Nassimbene and Donald G. Wooden, "Growth of Business Capital
Equipment, 1929-53," Survey of Current Business, XXXIV (December, 1954),
9See, for example, Goldsmith, A Study of Saving.
10Eric Schiff, "Gross Stocks Estimated from Past Installations,"
Review of Economics and Statistics, XL (May, 1958), 174-177; Myron J.
Gordon, 'The Valuation of Accounts at Current Cost," Accounting Review,
XXVIII (July, 1953), 373-384.
fraction of such assets appear on the market in any given time period.
Still, for reasons mentioned above in connection with reproducible
assets, current market value, whether actually sold or not, is the
appropriate measure--certainly more appropriate than original cost to
the owner, which is, after all, only the market price of some past per-
iod. Real property is usually appraised with greater ease than tangible
property in general; the problems here are more like problems of find-
ing data in general than tough theoretical problems.
Some interesting questions arise in connection with certain types
of public real property. Public domain in general presents no real prob-
lems, nor do military reservations nor, for that matter, parks and play-
grounds. We may appeal to the principle of opportunity cost for an
answer; if no opportunity cost, then no value on the national balance
sheet. It is true that some private properties have a greater market
value by virtue of their location adjacent to public properties, so that
any value in the public land is already reflected in surrounding private
land values. This, however, is a feature connon to private property as
well and cannot be advanced as an argument for excluding the value of
such land from the national wealth.11 As for the land underlying roads
and streets, it may be argued that there are not really any opportunity
costs since there is no real possibility of using such land for anything
else. As a practical matter it would appear not to make very much differ-
ence whether this land is included or excluded in the national wealth.
11This line of reasoning is used by Goldsmith in "Measuring Nation-
al Wealth in a System of Social Accounting" (loc. cit.), p. 48.
A rough estimate is included for the purposes of this study. Repro-
ducible assets like streets and sidewalks are a different matter, how-
ever; they represent resources which could well have been used else-
where and as such have a real opportunity cost.
As for intangible assets, the appropriate basis of valuation is,
in the writer's opinion, market value.12 Market prices are much more
readily available than in the case of tangible assets. Some assets are
effectively marketable at their face values--for example, currency and
demand deposits. Some interest-earning intangibles reflect changes in
interest rates in their market values; others respond by a change in their
own rates of return. Among the latter are time deposits and shares in
savings and loan associations. These may safely be carried at face
With respect to marketable federal securities, quotations are
readily available, while savings bonds may be valued on the basis of
their redemption schedules. As to corporate, foreign, state, and local
bonds, market quotations are available in many cases, and even those
privately-placed issues may be valued roughly on the basis of comparable
The proper valuation for bonds as liabilities presents another
problem. They are generally recorded on debtors' books at face value,
with any difference between this figure and the actual proceeds at the
12It might be reiterated at this point that intangible assets are
matters of importance only for economic units or for sectors; they do
not appear on the national balance sheet at all except in cases involv-
ing foreigners. The ensuing discussion, then, refers almost wholly to
sector balance sheets, not the national balance sheet itself,
time of issue shown separately as bond discount or premium. If bonds
are shown at market value from the standpoint of debtors, a discrepancy
will arise when we begin cancelling offsetting domestic assets and lia-
bilities. Problems of this nature are treated more fully in Chapter V
of this study, but this is essentially a valuation problem and will be
The method adopted here is to record bond liabilities at the
market value of the securities so that no discrepancy between the asset
and liability aspects of the same instrument appears. A superficial
reason is that the debtor could redeem any one of the bonds at the going
market value and hence discharge the liability.13 There is an even more
fundamental reason, however. Consider the case of a corporation which
has outstanding bonds due in ten years, bearing a 3 per cent coupon
rate. Suppose the bonds have a market price of 92, or an effective
yield of approximately 4 per cent. The market, in effect, is telling
the corporation that it must pay 4 per cent if it wishes to borrow; the
3 per cent coupons do not represent all the interest which the corporation
effectively pays in the future nor all of the interest which a present
purchaser of the instrument would receive.
In order to measure properly the present liability of the corpor-
ation and the future interest payments, the liability should be shown
at 92; then gradually increased over the ten-year period to 100, the
13It is true that the whole debt could probably not be discharged
without changing the market price. Valuation at the margin, however, is
a well established principle in economies,
annual addition to the liability being the same amount which the holder
would record as accruing effective interest. In the absence of further
changes in interest rates, market values would reflect precisely the
same valuations. Should interest rates again change, the resulting market
valuation is still the correct measuring rod for the liability. It will
be objected, of course, that in this example the corporation "owes" the
face amount, not 92 per cent of that amount. A debt maturing in ten
years, however, is a considerably different thing from a debt maturing
in two weeks or one year.
Something like this is actually done by business accountants in the
case of the issue of securities at other than face value (ignoring, for
the moment, flotation costs) which signifies that the coupon rate is not
the effective rate. The difference is recorded as discount or premium.
The gradual amortization of this quantity is a rough approximation of the
treatment suggested above; it deals with the difference between face and
market values on the date of issuance. We are merely extending this
same treatment to every date between issuance and redemption for which
a balance sheet is prepared.
Corporate stocks may also be shown as assets at market value since
a very large portion of all outstanding issues have some shares traded
at short intervals. The divergence between market values of stocks and
the underlying equities in assets presents a very serious problem in con-
solidation, which is discussed at some length in Chapter V.
As for intangible assets like patents, trademarks, and copyrights,
market value is again the theoretically correct method of valuation since
it represents the opportunity cost of retaining the advantage offered
by such assets. As we have suggested in Chapter I, these items are not
assets for the whole nation, but they are very definitely assets from
the viewpoint of the individual firm. This is one case in which the
concepts of private and national wealth diverge.14
As a practical matter, market values may rarely be available, so
that, for purposes of balance sheets for the business sectors, book
values (usually cost or nominal values) may be used. While technically
inconsistent, this procedure is probably the only possible course. Good-
will, another intangible, is closely bound up with the question of busi-
ness net worth versus market value of corporate stocks and will be dis-
cussed, as indicated above, in Chapter V.
Tangible assets in and claims against the United States owned by.
foreigners may be measured according to market valuations. With respect
to frequently traded securities, market values are easily obtained.
Wholly owned subsidiaries or others whose stocks seldom appear in markets
may be valued by means of underlying assets and liabilities. Directly
owned tangible assets may be valued in the same way as are correspond-
ing ones which are domestically owned. As already noted in Chapter I,
such assets should be included on the national balance sheet and then
14As noted in Chapter I, there are a few such cases which cannot
be satisfactorily resolved without doing violence either to sector wealth
on the one hand or national wealth on the other. The most important di-
vergence in valuation is that between corporate net worth and the market
value of outstanding stock, mentioned above. As also noted in Chapter I,
these instances are dealt with by showing the discrepancy in the national
balance sheet so that the user may resolve the matter in whichever way he
offset by a liability to foreigners in the same amount so that all
domestically situated wealth is covered.
American holdings of foreign securities may also be shown at
market values; many are traded regularly in American markets so that
direct dollar valuations are easily obtainable. Others are regularly
traded in foreign markets; values so determined may be converted at free
market rates or the same rates which are normally used for withdrawing
earnings. Direct investments in foreign countries, for which market
values are frequently not available, may be valued on the basis of orig-
inal costs adjusted to current costs, as is the case with domestic re-
All of the quantitatively important types of assets and liabili-
ties have been mentioned; there are others which might be considered in
a study devoted completely to such problems but which need not detain
SECTORING THE ECONOMY
Chapters I and II have examined the scope of the national balance
sheet and the methods of valuation which are suitable for use in its
preparation. There has been frequent reference to the sectors of the
economy, the idea of sectoring having been mentioned in the Introduction
to this study. It is the purpose of this chapter to discuss in more de-
tail the purposes of sectoring in social accounting and the type of
sectoring which is appropriate for the national balance sheet.
Every social accounting system which has set out to measure
national aggregates has--of necessity, it would seem--developed sector
information. The income and product accounts may be used as an example.
They purport to measure the nation's income and the nation's production,
yet in practice they do considerably more than this. Personal income,
one of the closely observed indicators of the nation's economic health,
is a sector concept, not a national concept, the sector in question
being persons or households. Sectoring is even more explicit in the
national savings-investment account. National investment (gross private
domestic investment plus net foreign investment) is, except for the
statistical discrepancy, numerically equal to personal saving plus
corporate saving (retained earnings) plus governments saving.
The basic reason for sectoring is that national aggregates in
themselves do not always convey enough information. Some disaggregation
adds considerably to the information about the economy which social
accounting statements can convey. This observation is particularly
relevant to the national balance sheet. It is informative, no doubt, to
know the total amount of the nation's wealth at any particular time. If
all items of national wealth were directly owned by individuals, a
simple statement of the national wealth might be sufficient, though it
would probably still be helpful to know something about its distribution
(a step in the direction of disaggregation).
It is characteristic of most developed economies, however, that
ownership is exercised largely-through claims, sometimes three cr four
stages removed from the actual possession of tangible items of healthh
which, except for foreign claims, comprise the nation's wealth. An indi-
vidual, for example, may have a claim against an insurance company which
has in turn a claim against a business firm exercised through the owner-
ship of corporate bonds. It is this structure of claims and liabilities
which makes necessary the inclusion of sector information in the nation-
al balance sheet. It is perhaps fair to say that the national balance
sheet, in throwing fresh light on the nature of this claims structure,
performs a mcre valuable service than in showing the total of national
The questions which arise in dividing the economy into sectors and
the combining or consolidating of economic units within the sectors must
be answered with a number of considerations in view.
The first consideration is the homogeneity of units combined; if
the information presented by the balance sheets is to have much
analytical usefulness, the economic units in a given sector should
respond similarly to changes in the economic variables. Complete homo-
geneity for each sector is probably a goal which cannot be wholly real-
ized, at least with a respectably small number of sectors. Within broad
limits, however, and by the judicious use of sub-sectors, it should be
possible to effect a reasonable compromise between the diversity of
units and the need for a readable statement. The reader may judge
whether this objective is subsequently accomplished.
A second consideration is the choice of an arrangement which will
show to best advantage the complex network of claims and counter-claims
through which America's wealth is owned. As a little reflection will
make clear, everything of value in the nation is ultimately owned by
someone--either individuals singly, groups of individuals in private
nonprofit organizations, or individuals collectively through their
governments. The larger share of the national wealth is owned indirect-
ly, that is, by businesses, which are in turn owned by individuals, or
even more remotely, through financial intermediaries such as banks, in-
surance companies, or investment trusts. To reduce to a single coherent
statement this vast, circuitous chain of assets and claims is one of the
primary purposes of the national balance sheet.
A third consideration is a form of organization which will facili-
tate the establishment of relationships with the other social accounting
statements, an objective briefly discussed above and analyzed in some
detail in Chapter VI of this study.
There are several different criteria by which economic units may
be grouped into sectors--by activity, for example, or by function, or on
the basis of the decision-making authority.1 Mining, ore-shipping, and
steel-making are three different activities from the viewpoint of input-
output accounting, but, if conducted by U.S. Steel, they are all treated
as a unit from the standpoint of decision-making authority. For another
example, certain assets and liabilities of the U.S. Treasury are tied to
the national money supply so that functionally they could be grouped
with the banking system under the heading "Money"; yet from the decision-
making standpoint banking is one sector and the U.S. Government is an-
The national income and products accounts are generally organized
around four sectors, assembled on the basis of activities. The sector
labelled "Persons" does all the consuming while "Business" does virtually
all the producing. Government, which would not fit into either cate-
gory, is shown separately, as are foreigners, so that the four familiar
sectors emerge. Housing, which is largely owned by individuals, is
nevertheless shown in the business sector, as are the commercial activ-
ities of government. Insofar as possible, any "productive" activity is
assigned to the business sector, and investment activity takes place
A discussion of the criteria used (differently) in national
income, input-output, and flow-of-funds accounts appears in a paper
by Stanley J. Sigel entitled "A Comparison of the Structures of Three
Social Accounting Systems" (Conference on Research in Income and
Wealth, Input-Output Analysis--an AppraisalrStudies in Income and
Wealth, Vol. XVIII (Princeton: Princeton University Press, 1955)] ).
This study is based on Dr. Sigel's Harvard Ph.D. dissertation.
Input-output systems, too, are oriented toward activity-sectoring,
but in quite a different way from national income.3 They expand in
great detail precisely those activities which are wholly suppressed in
income and product accounting, namely, the flows of intermediate goods
and services between "industries" which are the relevant sectors;
The Flow-of-Funds System,4 in its present stage of development, is
oriented toward sectoring on the basis of decision-making authority.
Since it expressly takes into account transactions in "financial" assets
and liabilities as well as in goods and services, this arrangement almost
necessarily follows. U.S. Steel, which for input-output purposes may be
allocated to several industries, is, for purposes of issuing securities,
paying dividends, borrowing from banks, and the like, one entity.
Of the three systems mentioned,5 the national balance sheet
2The latest published description of procedures and concepts in the
U.S. income ana product accounts is found in National Income, 1954 Edi-
tion (Iashington: US.. Government Printing Office, 1954) by the U.S. De-
partment of Commerce, Office of Business Economics.
3For a thorough summary of the state of input-output, see Confer-
ence on Research in Income and Wealth, Input-Output Analysis--an
Appraisal. The Technical Supplement gives the tables prepared in the 1947
inter-industry study as well as much detail on sources and methods.
The definitive published work to date is Flow of Funds in the
United States, 1939-1953 (Washington: Board of Governors of the Federal
Reserve System, 1955) by the Board of Governors of the Federal Reserve
System. Later mimeographed revisions and extensions are available on
5Technically speaking, there should be only one system of social
accounting from wh:-ch the various statements result. Tha fact that the
three above-mentioned have independently originated and as yet have not
been wholly integrated justifies the use of the term "three systems."
probably has most in common with the flow-of-funds accounts. Indeed,
the "financial" transactions shown therein are changes in balances of
intangible assets and liabilities, i.e., in accounts reflected in the
national balance sheet. In the absence of such a statement, the Flow-of-
Funds Section has constructed what amounts to the top half of a national
Like the flow-of-funds accounts, then, the balance sheet is orient-
ed toward sectoring by decision-making units. All decisions regarding
production, consumption, the distribution of asset holdings, the in-
currence of debt, investment and the like are made by some unit. They
are made, insofar as quantitative factors are concerned, on the basis of
the relevant flows (income, expenditure, etc.) and stocks (balance sheet
items), and on the quantitative terms on which the flows take place
(prices, wages, interest rates, etc.). Units which typically react in
a similar way are grouped together (as, for example, commercial banks
which would respond predictably to decreased reserves), and all of the
descriptive information about a given unit provided by a balance sheet
should be assigned to the same sector--that is, all of its assets and
liabilities should be shown within the same sector.6
6Income and product accounting, for example, considers the owner-
ship of one's home a business activity, so that the individual (consumer
sector) pays himself an imputed rent (business sector) in order to main-
tain, as far as possible, productive activity (shelter services here)
within the business sector. While this is a perfectly acceptable ab-
straction for deriving production and consumption totals, it is not well
suited to balance sheet objectives. Owner-occupied houses are assets
which compete with the remaining assets of their owners, and the mort-
gages against them are personal liabilities. Or, according to Graeme S.
Dorrance, split personalities should be avoided. These matters are dis-
A study of the evolution of the flow-of-funds work in the United
States may prove instructive, especially with respect to problems of
sectoring. The initial moneyflows project which was undertaken by
Professor Morris Copeland indicated the practicability of such a study,7
so that it was carried forward on an official basis by the Board of
Governors of the Federal Reserve System. The work is still in a state
of evolution, considerable revisions having been made since the 1955
publication already referred to in footnote four of this chapter.8 Some
of the shifts in sectoring which have been made serve to highlight the
problems which arise in this connection. Copeland's study divided the
economy into eleven sectors, which were rearranged and reduced to ten
in the Reserve Board's first published tables.9 Later revisions raised
the number to twelve; then reduced it to eleven. For purposes of com-
parison, the various sectoring schemes are presented in Table 2 below:
cussed in his very interesting unpublished paper "The Place of the
Balance Sheet in an Integrated System of Economic Accounts"
(Washington: International Monetary Fund, January 7, 1959 [Mimeo-
graphed] ), p. 7.
Morris A. Copeland, A Study of Moneyflows in the United States
(New York: National Bureau of Economic Research, 1952).
8Board of Governors of the Federal Reserve System, Flow of Funds
in the United States. 1939-1953.
COMPARATIVE METHODS OF SECTORING THE
ECONOMY IN VARIOUS FLOW-OF-FUNDS STUDIES
ships and Partner-
State and Local
Banks and Monetary
Security and Realty
Firms, et al.
Rest of the World
2. Corporate Business
3. Nonfarm Unincorpor-
4. Farm Business
5. Federal Government
6. State and Local
7. Banking System
9. Other Institutional
LO. Rest of the World
2. Corporate Business
3. Nonfarm Unincorpor-
4. Farm Business
5. Federal Government
6. State and Local
8. Savings Institu-
10. Finance, n.e.c.
11. Nonprofit Organ-
12. Rest of the World
*Combined with Consumers in more recent revisions.
Most units of the federal government are shown under that heading;
among the exceptions are the municipal government activities of Washington,
D.C., which are grouped with State and Local Governments, and the Ex-
change Stabilization Fund, which is grouped with the banking sector, as
are the Treasury monetary funds (gold, silver, currency, etc.). Postal
savings, shown under banking by Copeland and the original flow-of-funds
study, are moved back to the federal government in the latest revision.
Savings and loan associations, grouped by Copeland with Security and
Realty Finns, were treated as Other Institutional Investors in the flow-
of-funds study and then removed to the Savingo Institutions sector in a
later revision. Private pension funds are uniformly included in the In-
surance sector, and home ownership (owner-occupied) is treated as a
consumer function in contrast to national income, where it is regarded
as a business. Homes held for rental, assigned by Copeland to the
Security and Realty Firms category, are in the Reserve Board's study
treated as Unincorporated Businesses.
By way of contrast, Goldsmith's national balance sheet is confined
to seven sectors,10 namely, (1) Consumers, (2) Farm Business, (3) Non-
farm Non-corporate Business, (4) Financial Intermediaries, (5) Other
Corporate Business, (6) Federal Government, and (7) State and Local
Government. All of the studies include both farm and nonfarm households
under the heading of Consumers. Nonprofit organizations, included under
Consumers by Goldsmith, were originally treated by flew-of-funds as
Other Institutional Investors, then shown separately, and finally moved
to the Consumer sector. Federal trust funds and veterans' life insur-
ance reserves are classified as financial intermediaries by Goldsmith
but are shown under the federal government throughout the flow-of-funds
studies. Treasury monetary funds, on the other hand, are classified
10The latest published figures are for 1955, taken from the
National Bureau of Economic Research 37th Annual Report, p. 36, and
the Joint Economic Committee of the U.S. Congress has had them repro-
duced in The National Economic Accounts of the United States, p. 301.
with the federal government by Goldsmith, whereas flow-of-funds consid-
ers them part of the banking sector.
The foregoing brief summary of some alternative plans of sector-
ing will serve as background for a more detailed discussion here of
some of the issues involved. We shall take as a starting point the
minimal four-sector scheme used in income and product accounting and
work toward a more informative though not unwieldy statement. Our first
aim will be to dispose quickly of some of the more straight-forward,
relatively uncontroversial matters.
The government sector should undoubtedly be divided between the
federal government on the one hand and state and local units on the
other. The financial resources of the federal government are infin-
itely more flexible. Its cyclical behavior differs, especially since
it is formally committed to contra-cyclical action, and it is more
closely related to the nation's money supply. This treatment has been
followed consistently by Goldsmith and all of the flow-of-funds studies.
Another obvious need is the development of separate sectors for
financial activities. In the income and product accounts they are
shown as part of the business sector, an acceptable procedure in view
of the relatively small volumes of incomes and services which they orig-
inate. From the standpoint of funds flows or claims and liabilities,
however, they are vitally important. This fact is recognized in all of
the sectoring schemes reviewed above; it is stressed by Dorrance,11
11Dorrance, op. cit., p. 8.
Vandermeulen,12 Edey and Peacockl3--in fact, by virtually every writer
whose thinking has gone beyond income and product accounting.
A somewhat more difficult question arises as to the degree of
further division of financial institutions. At the very minimum it
would seem desirable to separate institutions whose liabilities are con-
sidered to be money from those that are not. This is done in all of the
various stages of the flow-of-funds work. As Dorrance points out, the
banking system, especially along with the central bank, is one of the
sectors which enjoys considerable independence of action; it may initi-
ate expansions or contractions in money and credit to which all of the
remaining asset-liability relationships in the economy must adjust.14
This, it may be noted, is the premise on which monetary policy is built.
A closely related question concerns the central bank: is it a
separate sector, or merely part of the banking sector? The flow-of-funds
group, oddly enough, completely consolidates the reserve banks into the
banking sector and suppresses all of their relationships with the com-
mercial banks. It is true that, quantitatively, the reserve banks hold
a small share of the national total of assets and liabilities; it is also
true that a subsector statement might be used to portray the relation-
ships. In view of the fact, however, that this is the locus of monetary
policy--in fact, one of the critical points at which coercion is brought
12Vandermeulen and Vandermeulen, op. cit., p. 323.
13Edey and Peacock, op. cit., p. 194.
14Dorrance, op. cit., p. 9.
to bear on the whole economy--it would seem appropriate to show the
reserve banks as a separate sector in the national balance sheet.
Furthermore, a consolidation at this point violates the principle of
homogeneity. The commercial banks and reserve banks are ordinarily
at cross purposes; it is the business of the reserve banks to "lean
against the wind," at least as the Federal Reserve System conceives
its own role. This opinion is shared by Edey and Peacock,15 and
Another issue is the treatment of the Treasury monetary
accounts--gold, gold certificates, silver, silver certificates, U.S.
notes, coin, the Exchange Stabilization Fund, and other minor mone-
tary assets and liabilities. These are consolidated with the bank-
ing sector in the flow-of-funds work; Goldsmith has left them, as
already noted, in the accounts of the federal government. Since we
are using as a guiding principle the exercise of control, it would
appear that such accounts belong in the federal government. True,
we are dividing the suppliers of money into several sectors, which
is what flow-of-funds has attempted to avoid with a consolidated
banking sector. Nevertheless, the monetary responsibilities are
divided in the United States, and our social accounting may as well
recognize as much. Furthermore, all of the external factors which
impinge upon the commercial banks should be highlighted in the nat-
ional balance sheet.
15Edey and Peacock, op. cit., p. 195.
16Dorrance, op. cit., p. 9.
The non-bank financial institutions are too heterogeneous a
group to avoid further sectoring. As indicated above, flow-of-
funds has tentatively settled on a thrae-fold division--savings in-
stitutions, insurance, and finance not elsewhere classified. Sav-
ings institutions are savings and loan associations, mutual savings
banks, and credit unions--units which offer their substantially simi-
lar liabilities as outlets for personal saving, and which quantita-
tively are sufficiently important to merit treatment as a separate
sector. The insurance sector includes life insurance companies
(stock and mutual), property insurance companies (stock and mutual),
fraternal order insurance plans, non-profit medical insurance plans,
and self-administcred pension activities. Publicly administered pen-
sion plans (federal, state, and local) are shown in the accounts of
the respective government units.
The remaining sector (finance not elsewhere classified) con-
tains a wide variety of units and is perhaps the one place where any-
thing resaabling homogeneity is unattainable. Fortunately, the annual
flows and year-end balances are not large; no great violence is done
to the accounts. Included at this point are open-end investment
companies; sales, industrial, and personal finance companies; mort-
gage companies; factors; and security and commodity dealers.
The flow-of-funds treatment outlined above offers few grounds
for objection. It would seem that closed-end investment companies
should be classified as financial instiLtuions, as should purely
financial holding companies where the various enterprises are not con-
solidated for tax or business accounting purposes. The flow-of-funds
group has informally expressed a similar view; difficulties of data,
and also of distinguishing between operating and holding companies,
account for the present treatment.
Drawing the proper lines between consumer and business activi-
ties presents a set of knotty problems as does the related question
of homogeneity of the resulting groups. First, to mention some of
1. Consumer and producer activities are sometimes closely
intertwined, especially for farmers, owners of unin-
corporated businesses, and owners of real property.
2. In principle, corporations belong in one sector and unin-
corporated businesses in another, since the line between
business and owner personal accounts is clear cut for
corporations. In practice, for small family-held corpor-
ations, the line between business and personal accounts
may be more blurred than for some partnerships or proprie-
3. If the consumer and producer activities of farmers and
entrepreneurs are divided (as is presently the case in
national income figures), a distinctly nonhomogeneous
consumer group results. This difficulty has been pointed
out by Friend,17 Klein,18 Margolis,19 and Morgan20--
to mention only a few.
4. "Business" or even "Corporate Business" is an extremely
5. If ownership of residences is a consumer activity, what
happens when an owner decides to rent his house to some-
one else? Or if ownership of rental residences is a per-
missable consumer activity, where do we draw the line?
May consumers own commercial real estate, or does this
make them businesses?
6. If sectoring is performed by dividing business into corpo-
rate and non-corporate, what happens when an individual
unit changes its legal form? If, for example, a partner-
ship incorporates, corporate assets and liabilities in the
national balance sheet would show a change not related to
current transactions at all.
7Friend, op. cit., p. 16; Irwin Friend and Irving B. Kravis,
"Entrepreneurial Income, Saving and Investment," American Economic
Review, XLVII (June, 1957), 271.
18Lawrence R. Klein and Julius Margolis, "Statistical Studies
of Unincorporated Business," Review of Economics and Statistics,
XXXVI (February, 1954), 33.
19Ibid.; Julius Margolis, "National Economic Accounting--a
Reorientation Needed," Review of Economics and Statistics, XXXIV
(November, 1952), 291.
20Morgan, "The Structure of Aggregate Personal Saving,"
loc. cit., p. 528.
With respect to the problems posed under section one above,
several solutions are possible. One is to treat both the producer
and consumer activities of farmers in the farm sector, and to treat
nonfarm unincorporated businesses and the consumer activities of their
owners together in a sector of their own. The remaining household
sector would exclude farmers and owners of unincorporated businesses
completely. This solution would have the very real advantage of
avoiding any artificial dissections of consumer assets and liabili-
ties. There is no rational way, in many cases, to divide a farmer's
or a dentist's bank account between his personal and business
Another proposal would consider proprietors to be dominated by
their consumer interests, so that their accounts would be shown in
the household sector.21 This solution, like the one above, avoids a
questionable split in the accounts of the individuals concerned.
Another solution lies in the present national income and flow-of-
funds practice; the division between consuming and producing functions
is arbitrarily made. Although farmers' residences have been conceded
to farm business activity, farm households are treated as consumers
like everyone else.
There is, under existing institutional arrangements in our
economy, probably no wholly satisfactory solution to these problems.
The writer inclines toward the solution adopted by national income and
21Dorrance, op. cit., p. 15.
flow-of-funds which maintains the distinction between business and
personal affairs. In the first place, nearly all partnerships as
well as many farms and other proprietorships do maintain records
which would permit a division to be made on a non-arbitrary basis,
and their number is probably increasing under the influence of the
Secondly, nearly everyone makes some decisions in his capacity
as a consumer as opposed to those he makes as a contributor to pro-
duction. True, the principle of maintaining all the accounts of a
single decision-maker in the same sector is violated, but the result-
ing split personality is not necessarily any more schizophrenic than
the salaried corporate executive who owns some stock in his enterprise.
We are really assuming that there are two decision-makers under one
hat, and we have kept the relevant accounts of each decision-maker in
the same sector.
Thirdly, the fact that some assets and liabilities must be
arbitrarily allocated is not necessarily fatal; errors are offset by
equal errors opposite in sign in the owner's proprietorship account.
If we erroneously allocate all of a farmer's $100,000 bank account to
his business, we have increased the net worth of the business and hence
the farmer's equity, which appears in his separate personal accounts.
While admittedly we have substituted a less liquid asset for money in
his personal accounts, we have disturbed neither his total assets,
total liabilities, nor net worth.
We might also meet the objection raised in item three above
(the nonhoaogeneity of a consumer sector containing farmers and other
entrepreneurs) by dividing the broadly inclusive consumer sector pro-
posed here into subsectors--farmers, entrepreneurs, and others. Each
of the consumer subsectors could be examined along with the appro-
priate business sector--farm households with farm business, and pro-
prietors with the unincorporated business sector.
With respect to the problem raised in item two above regarding
corporate and non-corporate business, it would probably be more in-
structive to divide businesses between "small" and "large" rather
than observe the legal distinctions between incorporated and unin-
corporated businesses. The term "large" is taken to mean those busi-
nesses whose shares are listed on stock exchanges or traded extensive-
ly over the counter. The term "small" refers to closely held busi-
nesses, be they corporations, partnerships, or proprietorships. This
distinction has recently been recognized in tax legislation, which
permits certain "small" corporations to be taxed as partnerships. The
same distinction would probably be made in social accounting, except
that the form of available data compels the present arrangement. In-
ternal Revenue Service tabulations are prepared regularly for corpo-
rations, though after a considerable time lag; they are prepared less
frequently for other returns. This adjustment, while desirable,
probably cannot be made at the present time.
The problem raised above regarding the nonhomogeneity of the
business sector has no satisfactory solution aside from preparing
supporting schedules to the main national balance sheet. Attempting
to insert even a limited number of major groups--such as railroads,
utilities, mining, etc.--ould probably result in an overly cumbersome
statement without doing much to improve the homogeneity of the sectors.
The best solution would seem to be showing three business sectors in
the national balance sheet--farms, unincorporated businesses, and
corporations. For the latter two sectors, complete subsectors could
be prepared showing all assets and liabilities for the major industry
groups. For physical assets--namely, inventories and fixed capital--
figures for the three main business sectors should be combined and
then redivided on a much finer basis--specifically, to conform to the
industry division employed in the input-output table so as to yield
capital coefficients for purposes of input-output analysis.22
As to the problems raised above regarding the ownership of real
property, this, in general, would appear to be a legitimate activity
for households. Real estate--whether residential, rental, or
commercial--competes with other earning household assets in the ab-
sorption of household savings; such transactions constitute an invest-
ment decision similar to the purchase of securities or other forms of
household wealth. Real estate holdings which go beyond the ownership
of property and the earnings of rentals--that is, properties used in
business enterprises also owned by the same household--should be
shown along with the other accounts of the business itself.
In the national income accounts, all ownership of real property
22See Conference on Research in Income and Wealth, Input-
Output Analysis--an Appraisal.
is considered to be a business activity. Flow-of-funds has assigned
owner-occupied residences to the consumer sector but treats all other
real property ownership as an unincorporated business activity.
Goldsmith draws the line at multi-family and commercial properties;
single residences--whether owner-occupied or rented--are treated as
assets of the household sector.23 Our treatment, which permits house-
holds to own all kinds of property directly, avoids split personalities
in all cases except those already split above with respect to unin-
The problem raised above regarding the shifting of units from one
sector to another is probably quantitatively unimportant. If shifts
of sufficient size should develop, they could be shown separately in
reconciling balance sheet figures from one year to the next. The
same is true of a similar problem--that of foreigners taking up perma-
nent residence in the United States or American citizens permanently
Private non-profit institutions--churches, schools, colleges,
foundations, labor, veterans, fraternal, and welfare groups, and the
like--present another problem in sectoring. National income treats
them as part of the personal sector; the early flow-of-funds treat-
ment was to classify them as "other institutional investors." Later
23U.S. Congress, Joint Economic Committee, Subcommittee on
Economic Statistics, op. cit., p. 302.
2See the discussion in Chapter I regarding the conceptual
boundaries of the national wealth.
shown by themselves as a separate sector, they have been more recent-
ly combined with the consumer sector in the latest revision, mostly
on grounds of incomplete data.
The national income treatment, suitable for purposes of measur-
ing income and product, should probably not be extended to other
social accounting measures.25 Though some of their activities might
be regarded as collective consumption, they are certainly not con-
sumers in the ordinary sense of the word nor do they belong in the con-
sumer sector from the standpoint of decision-making authority. While
some of their activities resemble those of businesses, the absence of
the profit motives and the relatively large role of transfer receipts
and expenditures preclude placing them in one of the business sectors.26
It is the belief of the writer that nonprofit organization holdings
of real estate and financial assets are sufficiently large to warrant
the establishment of a separate sector, at least for purposes of the
national balance sheet. Though their role in economic activity is
probably passive--that is, they react to rather than initiate changes--
it would be useful to have available sufficient information to find
25This treatment has not gone unchallenged with respect to the
resulting nonhomogeneity of the consumer sector. See, for example, S.
A. Goldberg and F. H. Leacy, "The National Accounts: Whither Now?"
Canadian Journal of Economics and Political Science, XXII (February,
1956), 73; Margolis, "National Economic Accounting--A Reorientation
Needed," loc. cit., p. 291.
26This is the treatment accorded them by Dorrance (op. cit.,
p. 19) after a discussion of some of the difficulties in classifying
them. The reason given is that showing them separately would
exaggerate their importance.
it is possible that their activities, especially construction,
represent a cyclically de-stabilizing factor. It may also be true
that their transfer receipts, treated as current "expenditures" by
the donors, are to some (and occasionally significant) extent saved.
Such transactions are labelled "personal saving" in the national in-
come accounts, which treatment, while consistent with the classifi-
cation of non-profit organizations, is hardly suitable for the
national balance sheet. The accumulated savings of such organizations
are not at the disposal of households in the formulation of their
consumption decisions, and this fact should be reflected in the
With respect to foreign transactions, it is technically un-
necessary to include any separate sector. While national income and
flow-of-funds both have a "Rest of the World" account, the nature of
the balance sheet is such that the appropriate information amarges in
any event. If all of the assets and liabilities of units within the
United States are consistently valued and then offset against each
other, claims against foreigners (assets) and claims of foreigners
against domestic units (liabilities) will remain unconsolidated and
will appear in the resulting statement of the national wealth (de-
fined as in Chaptcr I).
There are, however, numerous slips between the theoretical cup
and the technical lip, so that it would be advisable to include on a
memorandum basis a sector for foreigners. This procedure would also
facilitate relating the national balance sheet to the balance of
payment figures, as well as to net foreign investment in the income
and product accounts and the flow-of-funds foreign sector.
All of the major problems in sectoring have now been catalogued,
and the solutions thought to be appropriate to the national balance
sheet have been indicated. The resulting sectors, by way of surnary,
are indicated below together with some of the more important economic
units which are included in each.
PROPOSED SECTORS FOR THE NATIONAL BALANCE SHEET, WITH
MAJOR SECTOR COMPONENTS
Economic Units Included
2. Nonprofit Organizations
3. Farm Business
4. Nonfarm Unincorporated
5. Nonfarm Corporate
6. Federal Goverrment
7. State and Local
8. Federal Reserve System
9. Comercial Banks
10. Savings Institutions
12. Other Financial
13. Rest of the World
Household ownership of real estate, farm
households, and personal trust funds
Farm corporations and utility, marketing
and purchasing cooperatives
Trust Funds, Lending Agencies, Monetary
Funds, Enterprises and Insuring Agencies,
and Postal Savings
Trust Funds, Sinking Funds, Enterprises,
Semi-Independent Authorities, and the
Washington, D. C. Accounts
Savings and Loan Associations, Mutual Sav-
ings Banks, and Credit Unions
Life, Fire, and Other Property Insurers,
Insurance Activities of Nonprofit Organ-
izations, Self-Administered Pension Funds,
Investment Trusts, Finance Companies,
Mortgage Companies, Factors, Security
and Commodity Brokers
SAMPLE NATIONAL BALANCE SHEETS FOR 1949 AND 1950
The preceding three chapters have developed the conceptual struc-
ture within which national and sector wealth may be measured. It is
the purpose of this chapter to present illustrative national balance
sheets for two recent dates, to comment on some of the salient
features of the balance sheets, and indicate briefly the sources of
As we have already noted, a balance sheet prepared only for the
nation as a whole is a good deal less informative than one which con-
veys information about the various sectors of the nation as well.
Accordingly, the sample balance sheets presented below show each of
the thirteen sectors (as discussed in the preceding chapter) side by
side so that comparisons are easily possible and national totals are
readily obtained. The assets and liabilities shown are selected for
inclusion on the basis of the discussion in Chapter I; valuation of
the assets and liabilities is accomplished on the bases described in
The balance sheets are arranged so that both sector and national
wealth are shown, as, more important, is the way in which national and
sector wealth are related. Both the asset and liability aspects of
financial instruments are shown on the same line in order to facilitate
the process of consolidation. Demand deposits and currency, for ex-
ample, appear as assets of all of the sectors and, on the same line,
as liabilities of the federal government, Federal Reserve, and
commercial banking sectors.
National totals of assets and liabilities are shown in the col-
umn headed "Combined National Balance Sheet." Discrepancies and ad-
justments (which are discussed in some detail in the following chapter)
are entered in the next column. Financial assets and liabilities as
adjusted are then cancelled against each other so that only assets and
liabilities involving foreigners remain on the consolidated national
Assets are divided into tangible and intangible groupings on the
balance sheet, except that monetary metals, while tangible, are shown
along with intangible assets. Listing of assets is roughly in the
order of liquidity, though some complications are added by the scheme
of presentation adopted wherein both the asset and liability features
of a given financial instrument are treated in the same line of the bal-
ance sheet. For example, corporate stocks when viewed as assets are a
good deal more liquid than the corresponding net worth, at least with
respect to those stocks traded in organized markets. Long-term bonds
likewise may be considerably more liquid as assets than as liabilities,
since any individual bondholder may liquidate his holdings through the
bond market long before the issuer is obliged to make payment.
For financial assets and liabilities most of the balance sheet
figures are taken from unpublished materials made available by the Flow-
of-Funds Section at the Board of Governors.1 For tangible assets the
basic source is Goldsmith's savings study.2 Figures for tangible
assets were available only for 1949; gross investment less capital
consumption provided the figures for 1950, as explained below. Those
cases in which other sources were used or in which adjustments were
made to figures from the two basic sources are discussed in some detail
The flow-of-funds accounts structure consolidates into one sector
the Federal Reserve System, the commercial banks, and the Treasury
monetary activities. As explained in Chapter III, a great deal of in-
formation which is useful for the purposes of this study is thus
suppressed.. Accordingly, these three components have been separated;
Federal Reserve and commercial banks are shown as separate sectors,
while Treasury monetary activities have been transferred to the federal
government sector. Amounts for gold, gold certificates, other monetary
metals, treasury currency outstanding, and treasury cash holdings were
taken from the Federal Reserve Bulletin.3 The flow-of-funds study also
lBoard of Governors of the Federal Reserve System, "A Summary of
the Flow of Funds Accounts, 1950-1957" (Washington: Board of Govern-
ors of the Federal Reserve System, 1958 [Mimeographed]). These mat-
erials represent revisions and extensions of the previously published
Flow of Funds in the United States, 1939-1953.
Goldsmith, A Study of Saving, III, 57.
3Federal Reserve Bulletin, XXXVI (February, 1950), 185-194;
ibid., XXXVII (February, 1951), 170-178.
subtracts holdings of federal obligations by trust funds and other
federal agencies from the total outstanding; these amounts have been
restored as assets of the federal government sector while the corres-
ponding liability has been increased in the same amount.
Holdings by corporations of stocks of other corporations are not
shown in the flow-of-funds figures; Goldsmith's estimate is used to
make this adjustment to the 1949 figures. For 1950, the same percent-
age increase in stock values shown in the flow-of-funds accounts was
applied to corporate stock holdings. Year-end balances of direct for-
eign investments are not shown in the flow-of-funds tables, though
annual changes are. Year-end amounts for 1949 and 1950 were taken
from the Survey of Current Business.4
Household equities in unincorporated businesses are not shown in
the flow-of-funds year-end figures though annual changes are given.
The addition of tangible assets in the national balance sheet to in-
tangible assets and liabilities given in the flow-of-funds accounts
provides sufficient information for a calculation of unincorporated
business net worth, which is also shown as a household asset.
With respect to saving through life insurance contracts, it has
been necessary to adjust the funds-flow figures. These figures reflect
only cash surrender values of policies, though present values of future
payments computed actuarially are somewhat larger. In this study an
Samuel Pizer and John B. Boddie, "International Investment Posi-
tion of the United States," Survey of Current Business, XXXIV (May,
attempt has been made to show the full amount of life insurance re-
serves as a household asset, of which only the portion represented by
cash surrender values may be considered a liquid asset.
Goldsmith's figures for saving through life insurance seem, on
the other hand, to be too large since they include retained earnings
and other reserves of insurance companies in household savings as
well as policy reserves.5 Goldsmith's figure for consumer assets in
life insurance is adjusted by subtracting an estimate of retained
earnings and reserves other than policy reserves. The adjustment is
estimated from Internal Revenue Service figures.6
Savings through privately-administered pension funds is omitted
from flow-of-funds year-end assets and liabilities. Household equi-
ties in this study are taken to be the full amount of pension fund
assets. This procedure is not really satisfactory since fund assets
frequently fall far short of pension liabilities computed on an
actuarial basis. For newly established funds, past service credits
may not be fully funded for years. Even in the case of established
funds, unexpected wage and salary increases raise the basis on which
pension benefits are calculated and hence render past accumulations
inadequate. A fully informative national balance sheet would show the
full present values of actuarially computed pension liabilities as
5Goldsmith, A Study of Saving, II, 268.
6U.S. Treasury Department, Internal Revenue Service, Statistics
of Income for 1950 (Washington: U.S. Government Printing Office, 1954),
Part II, 126.
household assets and pension fund liabilities. Any fund liability in
excess of fund assets is, of course, a liability of the sponsoring
business enterprise, not only of the pension fund itself. Unfor-
tunately, the writer has not been able to discover an estimate of such
liabilities and is not sufficiently well informed in the specialized
field of insurance to make even a crude guess. The figure is undoubt-
edly a large one--sufficiently large to modify significantly the whole
structure of national assets and liabilities.
It has not been possible fully to reflect accrued, prepaid, and
deferred items in the national balance sheet. They have already been
included to an undeterminable extent in flow-of-funds trade credit
figures. The figures shown in the national balance sheet are taken
from Goldsmith's estimates of accured taxes, which are not included
in flow-of-funds figures.
Estimates of year-end tangible asset values have been taken from
the national balance sheet prepared by Goldsmith in connection with
his savings study. His method of computing tangible assets was to
cumulate annual investment expenditures, adjusted to current cost and
depreciated on the basis of current cost.
Since Goldsmith's work is sectored on a different basis than that
employed here, some changes in sectoring were necessary. Among these
were moving his government corporations sector into the federal govern-
ment accounts and dividing up his financial intermediaries sector into
the savings, insurance, and other finance sectors used in this study.
Residential property--which in Goldsmith's balance sheet appears
only in the household, farm, and corporate sectors--was reallocated
in a very rough fashion to unincorporated business, government--fed-
eral, state, and local--and insurance. The allocation from households
to business and government was made on the basis of the respective rent
receipts for 1950 (cash and imputed) given in National Income, 1954
Edition. The relatively small amounts allocated to the government and
insurance sectors are purely arbitrary.
Nonresidential property was reallocated in part from the busi-
ness sectors to households, again on the basis of rents received by
households from commercial and industrial property as reported for the
year 1950 in National Income, 1954 Edition. A small amount of pro-
ducer durables was arbitrarily allocated to banks and insurance
For farm assets and liabilities, flow-of-funds and Goldsmith's
figures were used even though figures prepared by the Department of
Agriculture in the annual "Balance Sheet of Agriculture" were avail-
able. Sufficient data to reconcile the many differences were not avail-
able, so that the above stated sources were used to maintain con-
sistency throughout the balance sheet.
Tangible asset figures were available only for 1949, so that
1950 figures had to be computed. This operation was done on a rough
basis in three steps. First, implicit deflators for broad components
of gross national product were applied to 1949 balance sheet amounts
to adjust them to a 1950 cost basis. The deflators employed are given
in the following table.
IMPLICIT DEFLATORS FOR NATIONAL ASSETS
(Index Numbers, 1947.100)
Assets 1949 1950
Consumer Durables 105.1 105.1
Producer Durables 113.3 115.7
Nonfarm Residential Structures 109.2 113.8
Other Structures 112.0 113.9
Source: National Income, 1954 Edition, p. 217.
The next step was to add capital expenditures for 1950, and the
third step was that of computing capital consumption on a current cost
basis. For producer and consumer durables, the sie relationships be-
tween depreciation and net stocks (cost less accumulated depreciation
reserves) as calculated by Goldsmith in 1949 were used to compute
1950 depreciation. Depreciation percentages are properly applied to
assets gross of depreciation reserves, not net, of course. However,
gross asset figures are not available, so the method used will serve
as a crude approximation. Residences were arbitrarily depreciated at
2 per cent and other structures at 2-1/2 per cent. Since only net
values (net of depreciation reserves) were known, it was assumed that
accumulated reserves amounted to one half of gross values; rates of
4 per cent and 5 per cent were applied to net values of residences and
other structures respectively. These procedures are hardly defensible
for anything but rough approximations, but they do serve in this study
to produce figures not too far removed from reality--figures which
illustrate the conceptual structure under study.
With respect to land values, Goldsmith's estimates are used for
1949. The Department of Agriculture estimated that farm land increased
in value by 14 per cent from 1949 to 1950 so that this figure was used
in arriving at a 1950 valuation.7 All other land was increased in
value by 5 per cent--a purely arbitrary assumption but one which again
serves to illustrate the relationships involved. Livestock values for
1950 were taken from "The Balance Sheet of Agriculture."8 Business
inventories in 1950 were derived by adding net 1950 accumulations plus
inventory valuation adjustments to the 1949 figures. Farm inventories
apparently changed very little, and federal government inventories
were left at the same figure as in 1949 in the absence of data.
Sector net worth estimates in the national balance sheets are taken
as assets less liabilities. They are divided into corporate and non-
corporate net worth so that the former may be compared with market
values of outstanding equity securities. Most of the corporate net
worth is located in the nonfinancial nonfarm corporate business sector;
the remainder consists of corporate farms, banks, stock insurance
companies, and corporations in the "Other Financial Institutions"
7"The Balance Sheet of Agriculture," Federal Reserve Bulletin,
XXXVII (September, 1951), 1094.
8Ibid. Goldsmith apparently used this source for his 1949 live-
sector. Farm net worth is allocated to corporate farms on the basis
of respective sales of corporate and noncorporate farms.9 Bank net
worth is entirely corporate, while 1949 insurance net worth is arbi-
trarily assigned 20 per cent to corporations and 80 per cent to mutual
companies. 1950 corporate insurance net worth is derived from 1949 net
worth plus retained earnings plus the estimated corporate share of
capital gains. The 1949 share of corporations in the net worth of the
"Other Financial Institutions" is arbitrarily set at one-half. The
1950 share is based on retained earnings plus net issues of new stocks.
There are a number of points where the figures in the national
balance sheet are less than satisfactory in addition to those mention-
ed above. Domestic holdings of foreign bonds are combined in the flow-
of-funds figures with corporate bonds; it has not been possible to
show them separately by holder on the national balance sheet. Figures
for holdings of foreign stocks are entirely lacking since they are not
given in the flow-of-funds accounts. This and the omission of various
minor holdings of foreign assets accounts for the discrepancy (1949)
in U.S. assets abroad of $2.7 billion between the national balance
sheet and the Department of Commerce estimates.10 Even the Coamerce
Department figures are not wholly suitable for the purposes of this
study since direct foreign investments and some other foreign assets
9Six per cent, as noted in Board of Governors of the Federal Re-
serve System, Flow of Funds in the United States, 1939-1953, p. 106.
10Pizer and Boddie, "International Investment Position of the
United States," loc. cit., p. 9.
are included at book values to parent companies, not at market values
or adjusted original cost.
A very serious shortcoming of the balance sheet is that depreci-
able tangible assets are shown net of depreciation reserves only. The
average age of the various components of the nation's stock of capital
as indicated by the size of the reserves is a fact of great signifi-
cance for cycle theory. Also in this connection, the balance sheet
should be supplemented by detailed schedules for the major types of
tangible assets; durable producer goods, for example, should be broken
down into twenty or thirty major categories.
The estimates for corporate and government bonds are such that
nominal rather than market values have been used, though market values
are preferable for the purposes of this study as indicated in Chapter
II. Part of the federal securities are valued at cost to holders, thus
giving rise to a small discrepancy which has not been removed since
the sectors to which the adjustments should be applied are not known.
Estimates for the market values or book values of patents and copy-
rights are not available so that these items have not been shown on the
balance sheet. Closed-end investment trusts, which should be shown in
the "Other Financial Institutions" sector, were left by flow-of-funds
in the corporate sector. This deficiency has not been remedied in the
absence of suitable data.
The foregoing discussion has catalogued the principal weaknesses
which remain in the sample balance sheets presented at the end of this
chapter. There are numerous possibilities for supplementary information
to be shown in supporting schedules, some of which possibilities are
discussed in the following paragraphs. The necessary information for
the actual preparation of such schedules is, for the most part, not
Assets and liabilities of the household sector may be broken down
by income classes, by age of heads of households, by occupation, or by
other relevant distinguishing characteristics. Studies of this sort
would probably proceed by "blowing up" sample data such as that obtain-
ed by the Federal Reserve Survey of Consumer Finances. Some analysis
of this type may be found in Goldsmith's savings study.11
Another useful means of organizing the data would be a breakdown
of tangible assets by input-output sectors. There is little reason to
divide the entire balance sheet in this fashion, even if it were pos-
sible to do so. One firm which would clearly lie in one sector on
the national balance sheet might well be distributed among several
input-output sectors. A division of intangible assets and liabilities
is thus not possible. A division of tangible assets, however, is
feasible and would provide capital coefficients for each industry.
An age distribution of tangible assets would provide useful in-
formation which might, among other things, be used in the estimation
of future replacement expenditures. A breakdown of debt by year of
scheduled repayment would provide useful information about future fund
flows and aid considerably any attempts to measure the burden of debt
11Goldsmith, A Study of Saving, III, 102ff.
for each sector.
If one is interested in measuring the growth of national wealth,
it is necessary to use constant dollar estimates of national wealth.
Tangible assets may be deflated by price indexes so as to make meaning-
ful comparisons possible, at least within the limitations imposed by
the use of index numbers. Any objections which might be raised to
this procedure are equally applicable to constant dollar estimates of
gross national product. It is doubtful, however, whether the entire
balance sheet could be adjusted for price changes in any meaningful
sense. True, assets and liabilities fixed in dollars could be convert-
ed by means of some sort of index of general purchasing power--the
consumer price index or the implicit gross national product deflator,
perhaps. This leaves open the question of deflating financial assets
like common stocks which are not fixed in dollars. There is some
doubt, though, as to just what would be accomplished by such operations.
The national wealth, except for net foreign assets, is virtually iden-
tical to the total of tangible assets, which can be adjusted by means
of specific, not general, deflators. The only information added to
the national wealth statement by the national balance sheet is the
means by which the national wealth is owned; hence each sector's share
in the national wealth. But each sector's share in the national wealth
can be determined at any particular date in current values. If total
national wealth can be deflated by means of adjustments to the values
of tangible assets, and if each sector's share is already known, there
would appear to be little point in going through the mechanics of
deflating the entire balance sheet--especially in view of the peculiar
problems involved in deflating common stocks, the means through which
a large share of the national wealth is owned.
The next matter to be considered is the national balance sheet it-
self. That an individual investigator like Goldsmith has been able to
assemble from various sources sufficient information to prepare a
national balance sheet and that enough supplementary information is
available so that the substantially different form and content can be
presented in this study testify to the feasibility of the balance sheet
The reader may feel entitled at this point to some suggestions as
to the value for economic analysis of balance sheet information, an ex-
pectation fulfilled in some measure in Chapter VII. It is probably
not possible to foresee some of the uses to which national balance
sheet data might be put; this has certainly been the case with respect
to national income and product figures. Theoretical and empirical
work, for all the heat generated in defense of one as against the
other, frequently go hand in hand. That some chapters of economic
theory remain unwritten is suggested more than once in the course of
At this point the reader's attention is invited to the sample bal-
ance sheets which appear on the following pages. Perhaps the most
interesting feature of the statement as presented here is the manner
in which offsetting assets and claims are arranged so as to facilitate
the process of consolidation. This is the major change from the
national balance sheets developed by Goldsmith. It is on this pro-
cess of consolidation that attention is focused in the following